(upbeat music) >> Welcome back to Supercloud 2, which is an open industry collaboration between technologists, consultants, analysts, and of course, practitioners, to help shape the future of cloud. And at this event, one of the key areas we're exploring is the intersection of cloud and data, and how building value on top of hyperscale clouds and across clouds is evolving, a concept we call supercloud. And we're pleased to welcome Harvir Singh, who's the chief data architect and global head of data at Western Union. Harvir, it's good to see you again. Thanks for coming on the program. >> Thanks, David, it's always a pleasure to talk to you. >> So many things stand out from when we first met, and one of the most gripping for me was when you said to me, "When data moves, money moves." And that's the world we live in today, and really have for a long time. Money has moved as bits, and when it has to move, we want it to move quickly, securely, and in a governed manner. And the pressure to do so is only growing. So tell us how that trend is evolved over the past decade in the context of your industry generally, and Western Union, specifically. Look, I always say to people that we are probably the first ones to introduce digital currency around the world because, hey, somebody around the world needs money, we move data to make that happen. That trend has actually accelerated quite a bit. If you look at the last 10 years, and you look at all these payment companies, digital companies, credit card companies that have evolved, majority of them are working on the same principle. When data moves, money moves. When data is stale, the money goes away, right? I think that trend is continuing, and it's not just the trend is in this space, it's also continuing in other spaces, specifically around, you know, acquisition of customers, communication with customers. It's all becoming digital, and it's, at the end of the day, it's all data being moved from one place or another. At the end of the day, you're not seeing the customer, but you're looking at, you know, the data that he's consuming, and you're making actionable items on it, and be able to respond to what they need. So I think 10 years, it's really, really evolved. >> Hmm, you operate, Western Union operates in more than 200 countries, and you you have what I would call a pseudo federated organization. You're trying to standardize wherever possible on the infrastructure, and you're curating the tooling and doing the heavy lifting in the data stack, which of course lessens the burden on the developers and the line of business consumers, so my question is, in operating in 200 countries, how do you deal with all the diversity of laws and regulations across those regions? I know you're heavily involved in AWS, but AWS isn't everywhere, you still have some on-prem infrastructure. Can you paint a picture of, you know, what that looks like? >> Yeah, a few years ago , we were primarily mostly on-prem, and one of the biggest pain points has been managing that infrastructure around the world in those countries. Yes, we operate in 200 countries, but we don't have infrastructure in 200 countries, but we do have agent locations in 200 countries. United Nations says we only have like 183 are countries, but there are countries which, you know, declare themselves countries, and we are there as well because somebody wants to send money there, right? Somebody has an agent location down there as well. So that infrastructure is obviously very hard to manage and maintain. We have to comply by numerous laws, you know. And the last few years, specifically with GDPR, CCPA, data localization laws in different countries, it's been a challenge, right? And one of the things that we did a few years ago, we decided that we want to be in the business of helping our customers move money faster, security, and with complete trust in us. We don't want to be able to, we don't want to be in the business of managing infrastructure. And that's one of the reasons we started to, you know, migrate and move our journey to the cloud. AWS, obviously chosen first because of its, you know, first in the game, has more locations, and more data centers around the world where we operate. But we still have, you know, existing infrastructure, which is in some countries, which is still localized because AWS hasn't reached there, or we don't have a comparable provider there. We still manage those. And we have to comply by those laws. Our data privacy and our data localization tech stack is pretty good, I would say. We manage our data very well, we manage our customer data very well, but it comes with a lot of complexity. You know, we get a lot of requests from European Union, we get a lot of requests from Asia Pacific every pretty much on a weekly basis to explain, you know, how we are taking controls and putting measures in place to make sure that the data is secured and is in the right place. So it's a complex environment. We do have exposure to other clouds as well, like Google and Azure. And as much as we would love to be completely, you know, very, very hybrid kind of an organization, it's still at a stage where we are still very heavily focused on AWS yet, but at some point, you know, we would love to see a world which is not reliant on a single provider, but it's more a little bit more democratized, you know, as and when what I want to use, I should be able to use, and pay-per-use. And the concept started like that, but it's obviously it's now, again, there are like three big players in the market, and, you know, they're doing their own thing. Would love to see them come collaborate at some point. >> Yeah, wouldn't we all. I want to double-click on the whole multi-cloud strategy, but if I understand it correctly, and in a perfect world, everything on-premises would be in the cloud is, first of all, is that a correct statement? Is that nirvana for you or not necessarily? >> I would say it is nirvana for us, but I would also put a caveat, is it's very tricky because from a regulatory perspective, we are a regulated entity in many countries. The regulators would want to see some control if something happens with a relationship with AWS in one country, or with Google in another country, and it keeps happening, right? For example, Russia was a good example where we had to switch things off. We should be able to do that. But if let's say somewhere in Asia, this country decides that they don't want to partner with AWS, and majority of our stuff is on AWS, where do I go from there? So we have to have some level of confidence in our own infrastructure, so we do maintain some to be able to fail back into and move things it needs to be. So it's a tricky question. Yes, it's nirvana state that I don't have to manage infrastructure, but I think it's far less practical than it said. We will still own something that we call it our own where we have complete control, being a financial entity. >> And so do you try to, I'm sure you do, standardize between all the different on-premise, and in this case, the AWS cloud or maybe even other clouds. How do you do that? Do you work with, you know, different vendors at the various places of the stack to try to do that? Some of the vendors, you know, like a Snowflake is only in the cloud. You know, others, you know, whether it's whatever, analytics, or storage, or database, might be hybrid. What's your strategy with regard to creating as common an experience as possible between your on-prem and your clouds? >> You asked a question which I asked when I joined as well, right? Which question, this is one of the most important questions is how soon when I fail back, if I need to fail back? And how quickly can I, because not everything that is sitting on the cloud is comparable to on-prem or is backward compatible. And the reason I say backward compatible is, you know, there are, our on-prem cloud is obviously behind. We haven't taken enough time to kind of put it to a state where, because we started to migrate and now we have access to infrastructure on the cloud, most of the new things are being built there. But for critical application, I would say we have chronology that could be used to move back if need to be. So, you know, technologies like Couchbase, technologies like PostgreSQL, technologies like Db2, et cetera. We still have and maintain a fairly large portion of it on-prem where critical applications could potentially be serviced. We'll give you one example. We use Neo4j very heavily for our AML use cases. And that's an important one because if Neo4j on the cloud goes down, and it's happened in the past, again, even with three clusters, having all three clusters going down with a DR, we still need some accessibility of that because that's one of the biggest, you know, fraud and risk application it supports. So we do still maintain some comparable technology. Snowflake is an odd one. It's obviously there is none on-prem. But then, you know, Snowflake, I also feel it's more analytical based technology, not a transactional-based technology, at least in our ecosystem. So for me to replicate that, yes, it'll probably take time, but I can live with that. But my business will not stop because our transactional applications can potentially move over if need to. >> Yeah, and of course, you know, all these big market cap companies, so the Snowflake or Databricks, which is not public yet, but they've got big aspirations. And so, you know, we've seen things like Snowflake do a deal with Dell for on-prem object store. I think they do the same thing with Pure. And so over time, you see, Mongo, you know, extending its estate. And so over time all these things are coming together. I want to step out of this conversation for a second. I just ask you, given the current macroeconomic climate, what are the priorities? You know, obviously, people are, CIOs are tapping the breaks on spending, we've reported on that, but what is it? Is it security? Is it analytics? Is it modernization of the on-prem stack, which you were saying a little bit behind. Where are the priorities today given the economic headwinds? >> So the most important priority right now is growing the business, I would say. It's a different, I know this is more, this is not a very techy or a tech answer that, you know, you would expect, but it's growing the business. We want to acquire more customers and be able to service them as best needed. So the majority of our investment is going in the space where tech can support that initiative. During our earnings call, we released the new pillars of our organization where we will focus on, you know, omnichannel digital experience, and then one experience for customer, whether it's retail, whether it's digital. We want to open up our own experience stores, et cetera. So we are investing in technology where it's going to support those pillars. But the spend is in a way that we are obviously taking away from the things that do not support those. So it's, I would say it's flat for us. We are not like in heavily investing or aggressively increasing our tech budget, but it's more like, hey, switch this off because it doesn't make us money, but now switch this on because this is going to support what we can do with money, right? So that's kind of where we are heading towards. So it's not not driven by technology, but it's driven by business and how it supports our customers and our ability to compete in the market. >> You know, I think Harvir, that's consistent with what we heard in some other work that we've done, our ETR partner who does these types of surveys. We're hearing the same thing, is that, you know, we might not be spending on modernizing our on-prem stack. Yeah, we want to get to the cloud at some point and modernize that. But if it supports revenue, you know, we'll invest in that, and get the, you know, instant ROI. I want to ask you about, you know, this concept of supercloud, this abstracted layer of value on top of hyperscale infrastructure, and maybe on-prem. But we were talking about the integration, for instance, between Snowflake and Salesforce, where you got different data sources and you were explaining that you had great interest in being able to, you know, have a kind of, I'll say seamless, sorry, I know it's an overused word, but integration between the data sources and those two different platforms. Can you explain that and why that's attractive to you? >> Yeah, I'm a big supporter of action where the data is, right? Because the minute you start to move, things are already lost in translation. The time is lost, you can't get to it fast enough. So if, for example, for us, Snowflake, Salesforce, is our actionable platform where we action, we send marketing campaigns, we send customer communication via SMS, in app, as well as via email. Now, we would like to be able to interact with our customers pretty much on a, I would say near real time, but the concept of real time doesn't work well with me because I always feel that if you're observing something, it's not real time, it's already happened. But how soon can I react? That's the question. And given that I have to move that data all the way from our, let's say, engagement platforms like Adobe, and particles of the world into Snowflake first, and then do my modeling in some way, and be able to then put it back into Salesforce, it takes time. Yes, you know, I can do it in a few hours, but that few hours makes a lot of difference. Somebody sitting on my website, you know, couldn't find something, walked away, how soon do you think he will lose interest? Three hours, four hours, he'll probably gone, he will never come back. I think if I can react to that as fast as possible without too much data movement, I think that's a lot of good benefit that this kind of integration will bring. Yes, I can potentially take data directly into Salesforce, but I then now have two copies of data, which is, again, something that I'm not a big (indistinct) of. Let's keep the source of the data simple, clean, and a single source. I think this kind of integration will help a lot if the actions can be brought very close to where the data resides. >> Thank you for that. And so, you know, it's funny, we sometimes try to define real time as before you lose the customer, so that's kind of real time. But I want to come back to this idea of governed data sharing. You mentioned some other clouds, a little bit of Azure, a little bit of Google. In a world where, let's say you go more aggressively, and we know that for instance, if you want to use Google's AI tools, you got to use BigQuery. You know, today, anyway, they're not sort of so friendly with Snowflake, maybe different for the AWS, maybe Microsoft's going to be different as well. But in an ideal world, what I'm hearing is you want to keep the data in place. You don't want to move the data. Moving data is expensive, making copies is badness. It's expensive, and it's also, you know, changes the state, right? So you got governance issues. So this idea of supercloud is that you can leave the data in place and actually have a common experience across clouds. Let's just say, let's assume for a minute Google kind of wakes up, my words, not yours, and says, "Hey, maybe, you know what, partnering with a Snowflake or a Databricks is better for our business. It's better for the customers," how would that affect your business and the value that you can bring to your customers? >> Again, I would say that would be the nirvana state that, you know, we want to get to. Because I would say not everyone's perfect. They have great engineers and great products that they're developing, but that's where they compete as well, right? I would like to use the best of breed as much as possible. And I've been a person who has done this in the past as well. I've used, you know, tools to integrate. And the reason why this integration has worked is primarily because sometimes you do pick the best thing for that job. And Google's AI products are definitely doing really well, but, you know, that accessibility, if it's a problem, then I really can't depend on them, right? I would love to move some of that down there, but they have to make it possible for us. Azure is doing really, really good at investing, so I think they're a little bit more and more closer to getting to that state, and I know seeking our attention than Google at this point of time. But I think there will be a revelation moment because more and more people that I talk to like myself, they're also talking about the same thing. I'd like to be able to use Google's AdSense, I would like to be able to use Google's advertising platform, but you know what? I already have all this data, why do I need to move it? Can't they just go and access it? That question will keep haunting them (indistinct). >> You know, I think, obviously, Microsoft has always known, you know, understood ecosystems. I mean, AWS is nailing it, when you go to re:Invent, it's all about the ecosystem. And they think they realized they can make a lot more money, you know, together, than trying to have, and Google's got to figure that out. I think Google thinks, "All right, hey, we got to have the best tech." And that tech, they do have the great tech, and that's our competitive advantage. They got to wake up to the ecosystem and what's happening in the field and the go-to-market. I want to ask you about how you see data and cloud evolving in the future. You mentioned that things that are driving revenue are the priorities, and maybe you're already doing this today, but my question is, do you see a day when companies like yours are increasingly offering data and software services? You've been around for a long time as a company, you've got, you know, first party data, you've got proprietary knowledge, and maybe tooling that you've developed, and you're becoming more, you're already a technology company. Do you see someday pointing that at customers, or again, maybe you're doing it already, or is that not practical in your view? >> So data monetization has always been on the charts. The reason why it hasn't seen the light is regulatory pressure at this point of time. We are partnering up with certain agencies, again, you know, some pilots are happening to see the value of that and be able to offer that. But I think, you know, eventually, we'll get to a state where our, because we are trying to build accessible financial services, we will be in a state that we will be offering those to partners, which could then extended to their customers as well. So we are definitely exploring that. We are definitely exploring how to enrich our data with other data, and be able to complete a super set of data that can be used. Because frankly speaking, the data that we have is very interesting. We have trends of people migrating, we have trends of people migrating within the US, right? So if a new, let's say there's a new, like, I'll give you an example. Let's say New York City, I can tell you, at any given point of time, with my data, what is, you know, a dominant population in that area from migrant perspective. And if I see a change in that data, I can tell you where that is moving towards. I think it's going to be very interesting. We're a little bit, obviously, sometimes, you know, you're scared of sharing too much detail because there's too much data. So, but at the end of the day, I think at some point, we'll get to a state where we are confident that the data can be used for good. One simple example is, you know, pharmacies. They would love to get, you know, we've been talking to CVS and we are talking to Walgreens, and trying to figure out, if they would get access to this kind of data demographic information, what could they do be better? Because, you know, from a gene pool perspective, there are diseases and stuff that are very prevalent in one community versus the other. We could probably equip them with this information to be able to better, you know, let's say, staff their pharmacies or keep better inventory of products that could be used for the population in that area. Similarly, the likes of Walmarts and Krogers, they would like to have more, let's say, ethnic products in their aisles, right? How do you enable that? That data is primarily, I think we are the biggest source of that data. So we do take pride in it, but you know, with caution, we are obviously exploring that as well. >> My last question for you, Harvir, is I'm going to ask you to do a thought exercise. So in that vein, that whole monetization piece, imagine that now, Harvir, you are running a P&L that is going to monetize that data. And my question to you is a there's a business vector and a technology vector. So from a business standpoint, the more distribution channels you have, the better. So running on AWS cloud, partnering with Microsoft, partnering with Google, going to market with them, going to give you more revenue. Okay, so there's a motivation for multi-cloud or supercloud. That's indisputable. But from a technical standpoint, is there an advantage to running on multiple clouds or is that a disadvantage for you? >> It's, I would say it's a disadvantage because if my data is distributed, I have to combine it at some place. So the very first step that we had taken was obviously we brought in Snowflake. The reason, we wanted our analytical data and we want our historical data in the same place. So we are already there and ready to share. And we are actually participating in the data share, but in a private setting at the moment. So we are technically enabled to share, unless there is a significant, I would say, upside to moving that data to another cloud. I don't see any reason because I can enable anyone to come and get it from Snowflake. It's already enabled for us. >> Yeah, or if somehow, magically, several years down the road, some standard developed so you don't have to move the data. Maybe there's a new, Mogli is talking about a new data architecture, and, you know, that's probably years away, but, Harvir, you're an awesome guest. I love having you on, and really appreciate you participating in the program. >> I appreciate it. Thank you, and good luck (indistinct) >> Ah, thank you very much. This is Dave Vellante for John Furrier and the entire Cube community. Keep it right there for more great coverage from Supercloud 2. (uplifting music)

>>Hey everyone. Welcome back to Las Vegas. This is the Cube's live coverage of day. One of snowflake summit 22 fourth annual. We're very happy to be here. A lot of people here, Lisa Martin with Dave Valante, David's always great to be at these events with you, but me. This one is shot out of the cannon from day one, data, data, data, data. That's what you heard of here. First, we have two guests joining us next, please. Welcome Matthew Scalian. Who's an alumni of the cube CEO and founder of Matillion and Jer staying chief data architect and global head of data engineering from Western union. Welcome gentlemen. Thank >>You. Great to be here. >>We're gonna unpack the Western union story in a second. I love that, but Matthew, I wanted to start with you, give the audience who might not be familiar with Matillion an overview, your vision, your differentiators, your joint value statement with snowflake, >>Of course. Well, first of all, thank you for having me on the cube. Again, Matillion S mission is to make the world's data useful, and we do that by providing a technology platform that allows our customers to load transform, synchronize, and orchestrate data on the snowflake data cloud. And on, on the cloud in general, we've been doing that for a number of years. We're co headquartered in the UK and the us, hence my dat accents. And we work with all sorts of companies, commercial scale, large end enterprises, particularly including of course, I'm delighted to say our friends at Western union. So that's why we're here today. >>And we're gonna talk about that in a second, but I wanna understand what's new with the data integration platform from Matillion perspective, lots of stuff coming out, give us an overview. >>Yeah, of course, it's been a really busy year and it's great to be here at snowflake summit to be able to share some of what we've been working on. You know, the Matillion platform is all about making our customers as productive as possible in terms of time to value insight on that analytics, data science, AI projects, like get you to value faster. And so the more technology we can put in the platform and the easier we can make it to use, the better we can achieve that goal. So this year we've, we've shipped a product that we call MDL 2.0, that's enterprise focused, exquisitely, easy to use batch data pipelines. So customers can load data even more simply into the snowflake data cloud, very excitingly we've also launched Matillion CDC. And so this is an industry first cloud native writer, head log based change data capture. >>I haven't come up with a shorter way of saying that, but, and surprise customers need this technology and it's been around for years, but mostly pre-cloud technology. That's been repurposed for the cloud. And so Matillion has rebuilt that concept for the cloud. And we launched that earlier this year. And of course we've continued to build out the core Matillion ETL platform that today over a thousand joint snowflake Matillion customers use, including Western union, of course we've been adding features there such as universal connectivity. And so a challenge that all data integration vendors have is having the right connectors for their source systems. Universal connectivity allows you to connect to any source system without writing code point and click. We shape that as well. So it's been a busy year, >>Was really simple. Sorry. I love that. He said that and it also sounded great with your accent. I didn't wanna >>Thank you. Excellent. Javier, talk about your role at Western union in, in what you've seen in terms of the evolution of the, the data stack. >>So in the last few years, well, a little bit of Western union, a 70 or 170 year old company, pretty much everybody knows what Western union is, right? Driving an interesting synergy from what Matthew says, when data moves money moves, that's what we do when he moves the da, he moves the data. We move the money. That's the synergy between, you know, us and the organization that support us from data move perspective. So what I've seen in the last few years is obviously a shift towards the cloud, but, you know, within the cloud itself, obviously there's a lot of players as well. And we as customers have always been wishing to have a short, smaller footprint of data so that the movement becomes a little lesser. You know, interestingly enough, in this conference, I've heard some very interesting stuff, which kind of helping me to bring that footprint down to a manageable number, to be more governed, to be more, you know, effective in terms of delivering more end results for my customers as well. >>So Matillion has been a great partner for us from our cloud adoption perspective. During the COVID times, we were a re we are a, you know, multi-channel organization. We have retail stores as well, our digital presence, but people just couldn't go to the retail stores. So we had to find ways to accelerate our adoption, make sure our systems are scaling and making sure that we are delivering the same experience to our customers. And that's where, you know, tools like Matillion came in and really, really partnered up with us to kind of bring it up to the level. >>So talk specifically about the stack evolution. Cause I have this sort of theory that everybody talks about injecting data and, and machine intelligence and AI and machine learning into apps. But the application development stack is like totally separate from the, the data analytics and the data pipeline stack. And the database is somewhere over here as well. How is that evolving? Are those worlds coming together? >>Some part of those words are coming together, but where I still see the difference is your heavy lifting will still happen on the data stack. You cannot have that heavy lifting on the app because if once the apps becomes heavy, you'll have trouble communicating with, with, with the organizations. You know, you need to be as lean as possible in the front end and make sure things are curated. Things are available on demand as soon as possible. And that's why you see all these API driven applications are doing really, really well because they're delivering those results back to the, the leaner applications much faster. So I'm a big proponent of, yes, it can be hybrid, but the majority of the heavy lifting still needs to happen down at the data layer, which is where I think snowflake plays a really good role >>In APIs are the connective tissue >>APIs connections. Yes. >>Also I think, you know, in terms of the, the data stack, there's another parallel that you can draw from applications, right? So technology is when they're new, we tend to do things in a granular way. We write a lot of code. We do a lot of sticking of things together with plasters and sticky tape. And it's the purview of high end engineers and people enthusiastic about that to get started. Then the business starts to see the value in this stuff, and we need to move a lot faster. And technology solutions come in and this is what the, the data cloud is all about, right? The technology getting out of the way and allowing people to focus on higher order problems of innovating around analytics, data applications, AI, machine learning, you know, that's also where Matillion sit as well as other companies in this modern enterprise data stack is technology vendors are coming in allowing organizations to move faster and have high levels of productivity. So I think that's a good parallel to application development. >>And's just follow up on that. When you think about data prep and you know, all the focus on data quality, you've got a data team, you know, in the data pipeline, a very specialized, maybe even hyper specialized data engineers, quality engineers, data, quality engineers, data analysts, data scientist, but they, and they serve a lot of different business lines. They don't necessarily have the business, they don't have the business context typically. So it's kind of this back and forth. Do you see that changing in your organization or, or the are the lines of business taking more responsibility for the data and, and addressing that problem? It's, >>It's like you die by thousand paper cuts or you just die. Right? That's the kind >>Of, right, >>Because if I say it's, it's good to be federated, it comes with its own flaws. But if I say, if it's good to be decentralized, then I'm the, the guy to choke, right? And in my role, I'm the guy to choke. So I've selectively tried to be a pseudo federated organization, where do I do have folks reporting into our organization, but they sit close to the line of business because the business understands data better. We are working with them hand in glove. We have dedicated teams that support them. And our problem is we are also regional. We are 200 countries. So the regional needs are very different than our us needs. Majority of the organizations that you probably end up talking to have like very us focused, 50 per more than 50% of our revenue is international. So we do, we are dealing with people who are international, their needs for data, their needs for quality and their needs for the, the delivery of those analytics and the data is completely different. And so we have to be a little bit more closer to the business than traditionally. Some, some organizations feel that they need >>To, is there need for the underlying infrastructure and the operational details that as diverse, or is that something that you bring standardization to the, >>So the best part about this, the cloud that happened to us is exactly that, because at one point of time, I had infrastructure in one country. I had another infrastructure sitting in another country, regional teams, making different different decisions of bringing in different tools. Now I can standardize. I will say, Matillion is our standard for doing ETL work. If this is the use case, but then it gets deployed across the geographies because the cloud helps us or the cloud platform helps us to manage it. Sitting down here. I have three centers around the world, you know, Costa Rica, India, and the us. I can manage 24 7 sitting here. No >>Problem. So the underlying our infrastructure is, is global, but the data needs are dealt with locally. Yep. >>One of the pav question, I was just thinking JVE is super well positioned funds for you, which is around that business domain knowledge versus technical expertise. Cause again, early in technology journeys tend, things tend to be very technical and therefore only high end engineers can do it, but high end engineers are scar. Right? Right. And, and also, I mean, we survey our hundreds of large enterprise customers and they tell us they spend two thirds of their time doing stuff they don't really want to do like reinventing the wheel, basic data movement and the low order staff. And so if you can make those people more productive and allow them to focus on higher value problems, but also bring pseudo technical people into it. Overall, the business can go a lot faster. And the way you do that is by making it easier. That's why Matillion is a low code NOCO platform, but Jer and Western union are doing this right. I >>Mean, I can't compete with AWS and Google to hire people. So I need to find people who are smart to figure the products that we have to make them work. I don't want them to spend time on infrastructure, Adam, I don't want them to spend time on trying to manage platforms. I want them to deliver the data, deliver the results to the business so that they can build and serve their customers better. So it's a little bit of a different approach, different mindset. I used to be in consulting for 17 years. I thought I knew it all, but it changed overnight when I own all of these systems. And I'm like, I need to be a little bit more smarter than this. I need to be more proactive and figure out what my business needs rather than what just from a technology needs. It's more what the business needs and how I can deliver that needs to them. So simple analogy, you know, I can build the best architecture in the world. It's gonna cost me an arm and leg, but I can't drive it because the pipeline is not there. So I can have a Ferrari, but I can't drive it. It's still capped at 80, 80 miles an hour. So rather than spend, rather than building one Ferrari, let me have 10 Toyotas or 10 Fs, which will go further along and do better for my cus my, for my customers. >>So how do you see this whole, we hearing about the data cloud. We hear about the marketplace, data products now, application development inside the data cloud. How do you see that affecting not so much the productivity of the data teams. I don't wanna necessarily say, but the product, the value that, that customers like you can get out >>Data. So data is moving closer to the business. That's the value I see, because you are injecting the business and you're injecting the application much more closer to the data because it, in the past, it was days and days of, you know, churn the data to actually clear results. Now the data has moved much closer. So I have a much faster turnaround time. The business can adapt and actually react much, much faster. It took us like 16 to 30 days to deliver, you know, data for marketing. Now I can turn it down in four hours. If I see something happening, I'll give you an example. The war in Ukraine happened. Let is shut down operations in Russia. Ukraine is cash swamp. There's no cash in Ukraine. We have cash. We roll out campaign, $0 money, transferred to Ukraine within four hours of the world going on. That's the impact that we have >>Massive impact. That's huge, especially with such a macro challenge going on, on the, in, in the world. Thank you so much for sharing the Matillion snowflake partnership story, how it's helping Western union really transform into a data company. We love hearing stories of organizations that are 170 years old that have always really been technology focused, but to see it come to life so quickly is pretty powerful. Guys. Thank you so much for your time. Thanks >>Guys. Thank you, having it. Thank >>You >>For Dave Velante and our guests. I'm Lisa Martin. You're watching the cubes live coverage of snowflake summit 22 live from Las Vegas. Stick around. We'll be back after a short break.

>> Announcer: From theCUBE Studios in Palo Alto and Boston, connecting with thought leaders all around the world, this is a Cube conversation. >> Hey welcome back everybody, Jeff Frick here with theCUBE. We are in our Palo Alto studios, COVID is still going on, so all of the interviews continue to be remote, but we're excited to have a Cube alumni, he hasn't been on for a long time, and this guy has been in the weeds of the storage industry for a very very long time and we're happy to have him on and get an update because there continues to be a lot of exciting developments. He's Phil Bullinger, he is the SVP and general manager, data center business unit from Western Digital joining us, I think for Colorado, so Phil, great to see you, how's the weather in Colorado today? >> Hi Jeff, it's great to be here. Well, it's a hot, dry summer here, I'm sure like a lot of places. But yeah, enjoying the summer through these unusual times. >> It is unusual times, but fortunately there's great things like the internet and heavy duty compute and store out there so we can get together this way. So let's jump into it. You've been in the he business a long time, you've been at Western Digital, you were at EMC, you worked on Isilon, and you were at storage companies before that. And you've seen kind of this never-ending up and to the right slope that we see kind of ad nauseum in terms of the amount of storage demands. It's not going anywhere but up, and please increase complexity in terms of unstructure data, sources of data, speed of data, you know the kind of classic big V's of big data. So I wonder, before we jump into specifics, if you can kind of share your perspective 'cause you've been kind of sitting in the Catford seat, and Western Digital's a really unique company; you not only have solutions, but you also have media that feeds other people solutions. So you guys are really seeing and ultimately all this compute's got to put this data somewhere, and a whole lot of it's sitting on Western Digital. >> Yeah, it's a great intro there. Yeah, it's been interesting, through my career, I've seen a lot of advances in storage technology. Speeds and feeds like we often say, but the advancement through mechanical innovation, electrical innovation, chemistry, physics, just the relentless growth of data has been driven in many ways by the relentless acceleration and innovation of our ability to store that data, and that's been a very virtuous cycle through what, for me, has been 30 years in enterprise storage. There are some really interesting changes going on though I think. If you think about it, in a relatively short amount of time, data has gone from this artifact of our digital lives to the very engine that's driving the global economy. Our jobs, our relationships, our health, our security, they all kind of depend on data now, and for most companies, kind of irrespective of size, how you use data, how you store it, how you monetize it, how you use it to make better decisions to improve products and services, it becomes not just a matter of whether your company's going to thrive or not, but in many industries, it's almost an existential question; is your company going to be around in the future, and it depends on how well you're using data. So this drive to capitalize on the value of data is pretty significant. >> It's a really interesting topic, we've had a number of conversations around trying to get a book value of data, if you will, and I think there's a lot of conversations, whether it's accounting kind of way, or finance, or kind of good will of how do you value this data? But I think we see it intrinsically in a lot of the big companies that are really data based, like the Facebooks and the Amazons and the Netflixes and the Googles, and those types of companies where it's really easy to see, and if you see the valuation that they have, compared to their book value of assets, it's really baked into there. So it's fundamental to going forward, and then we have this thing called COVID hit, which I'm sure you've seen all the memes on social media. What drove your digital transformation, the CEO, the CMO, the board, or COVID-19? And it became this light switch moment where your opportunities to think about it are no more; you've got to jump in with both feet, and it's really interesting to your point that it's the ability to store this and think about it now differently as an asset driving business value versus a cost that IT has to accommodate to put this stuff somewhere, so it's a really different kind of a mind shift and really changes the investment equation for companies like Western Digital about how people should invest in higher performance and higher capacity and more unified and kind of democratizing the accessibility that data, to a much greater set of people with tools that can now start making much more business line and in-line decisions than just the data scientist kind of on Mahogany Row. >> Yeah, as you mentioned, Jeff, here at Western Digital, we have such a unique kind of perch in the industry to see all the dynamics in the OEM space and the hyperscale space and the channel, really across all the global economies about this growth of data. I have worked at several companies and have been familiar with what I would have called big data projects and fleets in the past. But at Western Digital, you have to move the decimal point quite a few digits to the right to get the perspective that we have on just the volume of data that the world has just relentless insatiably consuming. Just a couple examples, for our drive projects we're working on now, our capacity enterprise drive projects, you know, we used to do business case analysis and look at their lifecycle capacities and we measured them in exabytes, and not anymore, now we're talking about zettabyte, we're actually measuring capacity enterprise drive families in terms of how many zettabyte they're going to ship in their lifecycle. If we look at just the consumption of this data, the last 12 months of industry TAM for capacity enterprise compared to the 12 months prior to that, that annual growth rate was north of 60%. And so it's rare to see industries that are growing at that pace. And so the world is just consuming immense amounts of data, and as you mentioned, the COVID dynamics have been both an accelerant in some areas, as well as headwinds in others, but it's certainly accelerated digital transformation. I think a lot of companies we're talking about, digital transformation and hybrid models and COVID has really accelerated that, and it's certainly driving, continues to drive just this relentless need to store and access and take advantage of data. >> Yeah, well Phil, in advance of this interview, I pulled up the old chart with all the different bytes, kilobytes, megabytes, gigabytes, terabytes, petabytes, exabytes, and zettabytes, and just per the Wikipedia page, what is a zettabyte? It's as much information as there are grains of sand in all the world's beaches. For one zettabyte. You're talking about thinking in terms of those units, I mean, that is just mind boggling to think that that is the scale in which we're operating. >> It's really hard to get your head wrapped around a zettabyte of storage, and I think a lot of the industry thinks when we say zettabyte scale era, that it's just a buzz word, but I'm here to say it's a real thing. We're measuring projects in terms of zettabytes now. >> That's amazing. Well, let's jump into some of the technology. So I've been fortunate enough here at theCUBE to be there at a couple of major announcements along the way. We talked before we turned the cameras on, the helium announcement and having the hard drive sit in the fish bowl to get all types of interesting benefits from this less dense air that is helium versus oxygen. I was down at the Mammer and Hammer announcement, which was pretty interesting; big heavy technology moves there, to again, increase the capacity of the hard drive's base systems. You guys are doing a lot of stuff on RISC-V I know is an Open source project, so you guys have a lot of things happening, but now there's this new thing, this new thing called zonedd storage. So first off, before we get into it, why do we need zoned storage, and really what does it now bring to the table in terms of a capability? >> Yeah, great question, Jeff. So why now, right? Because I mentioned storage, I've been in storage for quite some time. In the last, let's just say in the last decade, we've seen the advent of the hyperscale model and certainly a whole nother explosion level of data and just the veracity with which they hyperscalers can create and consume and process and monetize data. And of course with that, has also come a lot of innovation, frankly, in the compute space around how to process that data and moving from what was just a general purpose CPU model to GPU's and DPU's and so we've seen a lot of innovation on that side, but frankly, in the storage side, we haven't seen much change at all in terms of how operating systems, applications, file systems, how they actually use the storage or communicate with the storage. And sure, we've seen advances in storage capacities; hard drives have gone from two to four, to eight, to 10 to 14, 16, and now our leading 18 and 20 terabyte hard drives. And similarly, on the SSD side, now we're dealing with the capacities of seven, and 15, and 30 terabytes. So things have gotten larger, as you expect. And some interfaces have improved, I think NVME, which we'll talk about, has been a nice advance in the industry; it's really now brought a very modern scalable low latency multi-threaded interface to a NAM flash to take advantage of the inherent performance of transistor based persistent storage. But really when you think about it, it hasn't changed a lot. But what has changed is workloads. One thing that definitely has evolved in the space of the last decade or so is this, the thing that's driving a lot of this explosion of data in the industry is around workloads that I would characterize as sequential in nature, they're serial, you can capture it in written. They also have a very consistent life cycle, so you would write them in a big chunk, you would read them maybe in smaller pieces, but the lifecycle of that data, we can treat more as a chunk of data, but the problem is applications, operating systems, vial systems continue to interface with storage using paradigms that are many decades old. The old 512 byte or even Forte, Sector size constructs were developed in the hard drive industry just as convenient paradigms to structure what is an unstructured sea of magnetic grains into something structured that can be used to store and access data. But the reality is when we talk about SSDs, structure really matters, and so what has changed in the industry is the workloads are driving very very fresh looks at how more intelligence can be applied to that application OS storage device interface to drive much greater efficiency. >> Right, so there's two things going on here that I want to drill down on. On one hand, you talked about kind of the introduction of NAND and flash, and treating it like you did, generically you did a regular hard drive. But you could get away and you could do some things because the interface wasn't taking full advantage of the speed that was capable in the NAND. But NVME has changed that, and now forced kind of getting rid of some of those inefficient processes that you could live with, so it's just kind of classic next level step up and capabilities. One is you get the better media, you just kind of plug it into the old way. Now actually you're starting to put in processes that take full advantage of the speed that that flash has. And I think obviously prices have come down dramatically since the first introduction, where before it was always kind of a clustered off or super high end, super low latency, super high value apps, it just continues to spread and proliferate throughout the data center. So what did NVME force you to think about in terms of maximizing the return on the NAND and flash? >> Yeah, NVME, which we've been involved in the standardization, I think it's been a very successful effort, but we have to remember NVME is about a decade old, or even more when the original work started around defining this interface, but it's been very successful. The NVME standard's body is very productive cross company effort, it's really driven a significant change, and what we see now is the rapid adoption of NVME in all of data center architectures, whether it's very large hyperscale to classic on prem enterprise to even smaller applications, it's just a very efficient interface mechanism for connecting SSDs into a server. So we continue to see evolution at NVME, which is great, and we'll talk about ZNS today as one of those evolutions. We're also very keenly interested in NVME protocol over fabrics, and so one of the things that Western Digital has been talking about a lot lately is incorporating NVME over fabrics as a mechanism for now connecting shared storage into multiple post architectures. We think this is a very attractive way to build shared storage architectures of the future that are scalable, that are composable, that really have a lot more agility with respect to rack level infrastructure and applying that infrastructure to applications. >> Right, now one thing that might strike some people as kind of counterintuitive is within the zoned storage in zoning off parts of the media, to think of the data also kind of in these big chunks, is it feels contrary to kind of atomization that we're seeing in the rest of the data center, right? So smaller units of compute, smaller units of store, so that you can assemble and disassemble them in different quantities as needed. So what was the special attribute that you had to think about and actually come back and provide a benefit in actually kind of re-chunking, if you will, in these zones versus trying to get as atomic as possible? >> Yeah, it's a great question, Jeff, and I think it's maybe not intuitive in terms of why zoned storage actually creates a more efficient storage paradigm when you're storing stuff essentially in larger blocks of data, but this is really where the intersection of structure and workload and sort of the nature of the data all come together. If you turn back the clock maybe four or five years when SMR hard drives host managers SMR hard drives first emerged on the scene. This was really taking advantage of the fact that the right head on a hard disk drive is larger than the read head, or the read head can be much smaller, and so the notion of overlapping or shingling the data on the drive, giving the read head a smaller target to read, but the writer a larger write pad to write the data could actually, what we found was it increases aerial density significantly. And so that was really the emergence of this notion of sequentially written larger blocks of data being actually much more efficiently stored when you think about physically how it's being stored. What's very new now and really gaining a lot of traction is the SSD corollary to SMR on the hard drive, on the SSD side, we had the ZNS specification, which is, very similarly where you'd divide up the name space of an SSD into fixed size zones, and those zones are written sequentially, but now those zones are intimately tied to the underlying physical architecture of the NAND itself; the dyes, the planes, the read pages, the erase pages. So that, in treating data as a block, you're actually eliminating a lot of the complexity and the work that an SSD has to do to emulate a legacy hard drive, and in doing so, you're increasing performance and endurance and the predictable performance of the device. >> I just love the way that you kind of twist the lens on the problem, and on one hand, by rule, just looking at my notes here, the zoned storage device is the ZSD's introduce a number of restrictions and limitations and rules that are outside the full capabilities of what you might do. But in doing so, an aggregate, the efficiency, and the performance of the system in the whole is much much better, even though when you first look at it, you think it's more of a limiter, but it's actually opens up. I wonder if there's any kind of performance stats you can share or any kind of empirical data just to give people kind of a feel for what that comes out as. >> So if you think about the potential of zoned storage in general and again, when I talk about zoned storage, there's two components; there's an HDD component of zoned storage that we refer to as SMR, and there's an SSD version of that that we call ZNS. So we think about SMR, the value proposition there is additional capacity. So effectively in the same drive architecture, with roughly the same bill of material used to build the drive, we can overlap or shingle the data on the drive. And generally for the customer, additional capacity. Today with our 18, 20 terabyte offerings that's on the order of just over 10%, but that delta is going to increase significantly going forward to 20% or more. And when you think about a hyperscale customer that has not hundreds or thousands of racks, but tens of thousands of racks. A 10 or 20% improvement in effective capacity is a tremendous TCO benefit, and the reason we do that is obvious. I mean, the economic paradigm that drives large at-scale data centers is total custom ownership, both acquisition costs and operating costs. And if you can put more storage in a square tile of data center space, you're going to generally use less power, you're going to run it more efficiently, you're actually, from an acquisition cost, you're getting a more efficient purchase of that capacity. And in doing that, our innovation, we benefit from it and our customers benefit from it. So the value proposition for zoned storage in capacity enterprise HDV is very clear, it's additional capacity. The exciting thing is, in the SSD side of things, or ZNS, it actually opens up even more value proposition for the customer. Because SSDs have had to emulate hard drives, there's been a lot of inefficiency and complexity inside an enterprise SSD dealing with things like garbage collection and right amplification reducing the endurance of the device. You have to over-provision, you have to insert as much as 20, 25, even 28% additional man bits inside the device just to allow for that extra space, that working space to deal with delete of data that are smaller than the block erase that the device supports. So you have to do a lot of reading and writing of data and cleaning up. It creates for a very complex environment. ZNS by mapping the zoned size with the physical structure of the SSD essentially eliminates garbage collection, it reduces over-provisioning by as much as 10x. And so if you were over provisioning by 20 or 25% on an enterprise SSD, and a ZNS SSD, that can be one or two percent. The other thing I have to keep in mind is enterprise SSD is typically incorporate D RAM and that D RAM is used to help manage all those dynamics that I just mentioned, but with a much simpler structure where the pointers to the data can be managed without all the D RAM. We can actually reduce the amount of D RAM in an enterprise SSD by as much as eight X. And if you think about the MILA material of an enterprise SSD, D RAM is number two on the list in terms of the most expensive bomb components. So ZNS and SSDs actually have a significant customer total cost of ownership impact. It's an exciting standard, and now that we have the standard ratified through the NVME working group, it can really accelerate the development of the software ecosystem around. >> Right, so let's shift gears and talk a little bit about less about the tech and more about the customers and the implementation of this. So you talked kind of generally, but are there certain types of workloads that you're seeing in the marketplace where this is a better fit or is it just really the big heavy lifts where they just need more and this is better? And then secondly, within these hyperscale companies, as well as just regular enterprises that are also seeing their data demands grow dramatically, are you seeing that this is a solution that they want to bring in for kind of the marginal kind of next data center, extension of their data center, or their next cloud region? Or are they doing lift and shift and ripping stuff out? Or do they enough data growth organically that there's plenty of new stuff that they can put in these new systems? >> Yeah, I love that. The large customers don't rip and shift; they ride their assets for a long lifecycle, 'cause with the relentless growth of data, you're primarily investing to handle what's coming in over the transom. But we're seeing solid adoption. And in SMRS you know we've been working on that for a number of years. We've got significant interest and investment, co-investment, our engineering, and our customer's engineering adapting the application environment's to take advantage of SMR. The great thing is now that we've got the NVME, the ZNS standard gratified now in the NVME working group, we've got a very similar, and all approved now, situation where we've got SMR standards that have been approved for some time, and the SATA and SCSI standards. Now we've got the same thing in the NVME standard, and the great thing is once a company goes through the lift, so to speak, to adapt an application, file system, operating system, ecosystem, to zoned storage, it pretty much works seamlessly between HDD and SSD, and so it's not an incremental investment when you're switching technologies. Obviously the early adopters of these technologies are going to be the large companies who design their own infrastructure, who have mega fleets of racks of infrastructure where these efficiencies really really make a difference in terms of how they can monetize that data, how they compete against the landscape of competitors they have. For companies that are totally reliant on kind of off the shelf standard applications, that adoption curve is going to be longer, of course, because there are some software changes that you need to adapt to enable zoned storage. One of the things Western Digital has done and taken the lead on is creating a landing page for the industry with zoned storage.io. It's a webpage that's actually an area where many companies can contribute Open source tools, code, validation environments, technical documentation. It's not a marketeering website, it's really a website built to land actual Open source content that companies can use and leverage and contribute to to accelerate the engineering work to adapt software stacks to zoned storage devices, and to share those things. >> Let me just follow up on that 'cause, again, you've been around for a while, and get your perspective on the power of Open source. And it used to be the best secrets, the best IP were closely guarded and held inside, and now really we're in an age where it's not necessarily. And the brilliant minds and use cases and people out there, just by definition, it's more groups of engineers, more engineers outside your building than inside your building, and how that's really changed kind of a strategy in terms of development when you can leverage Open source. >> Yeah, Open source clearly has accelerated innovation across the industry in so many ways, and it's the paradigm around which companies have built business models and innovated on top of it, I think it's always important as a company to understand what value ad you're bringing, and what value ad the customers want to pay for. What unmet needs in your customers are you trying to solve for, and what's the best mechanism to do that? And do you want to spend your RND recreating things, or leveraging what's available and innovating on top of it? It's all about ecosystem. I mean, the days where a single company could vertically integrate top to bottom a complete end solution, you know, those are fewer and far between. I think it's about collaboration and building ecosystems and operating within those. >> Yeah, it's such an interesting change, and one more thing, again, to get your perspective, you run the data center group, but there's this little thing happening out there that we see growing, IOT, in the industrial internet of things, and edge computing as we try to move more compute and store and power kind of outside the pristine world of the data center and out towards where this data is being collected and processed when you've got latency issues and all kinds of reasons to start to shift the balance of where the compute is and where the store and relies on the network. So when you look back from the storage perspective in your history in this industry and you start to see basically everything is now going to be connected, generating data, and a lot of it is even Opensource. I talked to somebody the other day doing kind of Opensource computer vision on surveillance video. So the amount of stuff coming off of these machines is growing in crazy ways. At the same time, it can't all be processed at the data center, it can't all be kind of shipped back and then have a decision and then ship that information back out to. So when you sit back and look at Edge from your kind of historical perspective, what goes through your mind, what gets you excited, what are some opportunities that you see that maybe the laymen is not paying close enough attention to? >> Yeah, it's really an exciting time in storage. I get asked that question from time to time, having been in storage for more than 30 years, you know, what was the most interesting time? And there's been a lot of them, but I wouldn't trade today's environment for any other in terms of just the velocity with which data is evolving and how it's being used and where it's being used. A TCO equation may describe what a data center looks like, but data locality will determine where it's located, and we're excited about the Edge opportunity. We see that as a pretty significant, meaningful part of the TAM as we look three to five years. Certainly 5G is driving much of that, I think just any time you speed up the speed of the connected fabric, you're going to increase storage and increase the processing the data. So the Edge opportunity is very interesting to us. We think a lot of it is driven by low latency work loads, so the concept of NVME is very appropriate for that, we think, in general SSDs deployed and Edge data centers defined as anywhere from a meter to a few kilometers from the source of the data. We think that's going to be a very strong paradigm. The workloads you mentioned, especially IOT, just machine-generated data in general, now I believe, has eclipsed human generated data, in terms of just the amount of data stored, and so we think that curve is just going to keep going in terms of machine generated data. Much of that data is so well suited for zoned storage because it's sequential, it's sequentially written, it's captured, and it has a very consistent and homogenous lifecycle associated with it. So we think what's going on with zoned storage in general and ZNS and SMR specifically are well suited for where a lot of the data growth is happening. And certainly we're going to see a lot of that at the Edge. >> Well, Phil, it's always great to talk to somebody who's been in the same industry for 30 years and is excited about today and the future. And as excited as they have been throughout their whole careers. So that really bodes well for you, bodes well for Western Digital, and we'll just keep hoping the smart people that you guys have over there, keep working on the software and the physics, and the mechanical engineering and keep moving this stuff along. It's really just amazing and just relentless. >> Yeah, it is relentless. What's exciting to me in particular, Jeff, is we've driven storage advancements largely through, as I said, a number of engineering disciplines, and those are still going to be important going forward, the chemistry, the physics, the electrical, the hardware capabilities. But I think as widely recognized in the industry, it's a diminishing curve. I mean, the amount of energy, the amount of engineering effort, investment, that cost and complexity of these products to get to that next capacity step is getting more difficult, not less. And so things like zoned storage, where we now bring intelligent data placement to this paradigm, is what I think makes this current juncture that we're at very exciting. >> Right, right, well, it's applied AI, right? Ultimately you're going to have more and more compute power driving the storage process and how that stuff is managed. As more cycles become available and they're cheaper, and ultimately compute gets cheaper and cheaper, as you said, you guys just keep finding new ways to move the curve in. And we didn't even get into the totally new material science, which is also coming down the pike at some point in time. >> Yeah, very exciting times. >> It's been great to catch up with you, I really enjoy the Western Digital story; I've been fortunate to sit in on a couple chapters, so again, congrats to you and we'll continue to watch and look forward to our next update. Hopefully it won't be another four years. >> Okay, thanks Jeff, I really appreciate the time. >> All right, thanks a lot. All right, he's Phil, I'm Jeff, you're watching theCUBE. Thanks for watching, we'll see you next time.

>> Live from Stanford University. It's the Cube covering global Women in Data Science conference brought to you by Silicon Angle media. >> Welcome back to the key. We air live at Stanford University for the fourth annual Women in Data Science Conference. The Cube has had the pleasure of being here all four years on I'm welcoming Back to the Cube, one of our distinguished alumni Janet George, the fellow chief data officer, scientists, big data and cognitive computing at Western Digital. Janet, it's great to see you. Thank you. Thank you so much. So I mentioned yes. Fourth, Annie will women in data science. And it's been, I think I met you here a couple of years ago, and we look at the impact. It had a chance to speak with Margo Garrett's in a about an hour ago, one of the co founders of Woods saying, We're expecting twenty thousand people to be engaging today with the Livestream. There are wigs events in one hundred and fifty locations this year, fifty plus countries expecting about one hundred thousand people to engage the attention. The focus that they have on data science and the opportunities that it has is really palpable. Tell us a little bit about Western Digital's continued sponsorship and what makes this important to you? >> So Western distal has recently transformed itself as a company, and we are a data driven company, so we are very much data infrastructure company, and I think that this momentum off A is phenomenal. It's just it's a foundational shift in the way we do business, and this foundational shift is just gaining tremendous momentum. Businesses are realizing that they're going to be in two categories the have and have not. And in order to be in the half category, you have started to embrace a You've got to start to embrace data. You've got to start to embrace scale and you've got to be in the transformation process. You have to transform yourself to put yourself in a competitive position. And that's why Vest Initial is here, where the leaders in storage worldwide and we'd like to be at the heart of their data is. >> So how has Western Digital transform? Because if we look at the evolution of a I and I know you're give you're on a panel tan, you're also giving a breakout on deep learning. But some of the importance it's not just the technical expertise. There's other really important skills. Communication, collaboration, empathy. How has Western digital transformed to really, I guess, maybe transform the human capital to be able to really become broad enough to be ableto tow harness. Aye, aye, for good. >> So we're not just a company that focuses on business for a We're doing a number of initiatives One of the initiatives were doing is a I for good, and we're doing data for good. This is related to working with the U. N. We've been focusing on trying to figure out how climate change the data that impacts climate change, collecting data and providing infrastructure to store massive amounts of species data in the environment that we've never actually collected before. So climate change is a huge area for us. Education is a huge area for us. Diversity is a huge area for us. We're using all of these areas as launching pad for data for good and trying to use data to better mankind and use a eye to better mankind. >> One of the things that is going on at this year's with second annual data fun. And when you talk about data for good, I think this year's Predictive Analytics Challenge was to look at satellite imagery to train the model to evaluate which images air likely tohave oil palm plantations. And we know that there's a tremendous social impact that palm oil and oil palm plantations in that can can impact, such as I think in Borneo and eighty percent reduction in the Oregon ten population. So it's interesting that they're also taking this opportunity to look at data for good. And how can they look at predictive Analytics to understand how to reduce deforestation like you talked about climate and the impact in the potential that a I and data for good have is astronomical? >> That's right. We could not build predictive models. We didn't have the data to put predictive boats predictive models. Now we have the data to put put out massively predictive models that can help us understand what change would look like twenty five years from now and then take corrective action. So we know carbon emissions are causing very significant damage to our environment. And there's something we can do about it. Data is helping us do that. We have the infrastructure, economies of scale. We can build massive platforms that can store this data, and then we can. Alan, it's the state at scale. We have enough technology now to adapt to our ecosystem, to look at disappearing grillers, you know, to look at disappearing insects, to look at just equal system that be living, how, how the ecosystem is going to survive and be better in the next ten years. There's a >> tremendous amount of power that data for good has, when often times whether the Cube is that technology conferences or events like this. The word trust issues yes, a lot in some pretty significant ways. And we often hear that data is not just the life blood of an organization, whether it's in just industry or academia. To have that trust is essential without it. That's right. No, go. >> That's right. So the data we have to be able to be discriminated. That's where the trust comes into factor, right? Because you can create a very good eh? I'm odder, or you can create a bad air more so a lot depends on who is creating the modern. The authorship of the model the creator of the modern is pretty significant to what the model actually does. Now we're getting a lot of this new area ofthe eyes coming in, which is the adversarial neural networks. And these areas are really just springing up because it can be creators to stop and block bad that's being done in the world next. So, for example, if you have malicious attacks on your website or hear militias, data collection on that data is being used against you. These adversarial networks and had built the trust in the data and in the so that is a whole new effort that has started in the latest world, which is >> critical because you mentioned everybody. I think, regardless of what generation you're in that's on. The planet today is aware of cybersecurity issues, whether it's H vac systems with DDOS attacks or it's ah baby boomer, who was part of the fifty million Facebook users whose data was used without their knowledge. It's becoming, I won't say accepted, but very much commonplace, Yes, so training the A I to be used for good is one thing. But I'm curious in terms of the potential that individuals have. What are your thoughts on some of these practices or concepts that we're hearing about data scientists taking something like a Hippocratic oath to start owning accountability for the data that they're working with. I'm just curious. What's >> more, I have a strong opinion on this because I think that data scientists are hugely responsible for what they are creating. We need a diversity of data scientists to have multiple models that are completely divorce, and we have to be very responsible when we start to create. Creators are by default, have to be responsible for their creation. Now where we get into tricky areas off, then you are the human auto or the creator ofthe Anay I model. And now the marshal has self created because it a self learned who owns the patent, who owns the copyright to those when I becomes the creator and whether it's malicious or non malicious right. And that's also ownership for the data scientist. So the group of people that are responsible for creating the environment, creating the morals the question comes into how do we protect the authors, the uses, the producers and the new creators off the original piece of art? Because at the end of the day, when you think about algorithms and I, it's just art its creation and you can use the creation for good or bad. And as the creation recreates itself like a learning on its own with massive amounts of data after an original data scientist has created the model well, how we how to be a confident. So that's a very interesting area that we haven't even touched upon because now the laws have to change. Policies have to change, but we can't stop innovation. Innovation has to go, and at the same time we have to be responsible about what we innovate >> and where do you think we are? Is a society in terms of catching As you mentioned, we can't. We have to continue innovation. Where are we A society and society and starting to understand the different principles of practices that have to be implemented in order for proper management of data, too. Enable innovation to continue at the pace that it needs. >> June. I would say that UK and other countries that kind of better than us, US is still catching up. But we're having great conversations. This is very important, right? We're debating the issues. We're coming together as a community. We're having so many discussions with experts. I'm sitting in so many panels contributing as an Aye aye expert in what we're creating. What? We see its scale when we deploy an aye aye, modern in production. What have we seen as the longevity of that? A marker in a business setting in a non business setting. How does the I perform and were now able to see sustained performance of the model? So let's say you deploy and am are in production. You're able inform yourself watching the sustained performance of that a model and how it is behaving, how it is learning how it's growing, what is its track record. And this knowledge is to come back and be part of discussions and part of being informed so we can change the regulations and be prepared for where this is going. Otherwise will be surprised. And I think that we have started a lot of discussions. The community's air coming together. The experts are coming together. So this is very good news. >> Theologian is's there? The moment of Edward is building. These conversations are happening. >> Yes, and policy makers are actively participating. This is very good for us because we don't want innovators to innovate without the participation of policymakers. We want the policymakers hand in hand with the innovators to lead the charter. So we have the checks and balances in place, and we feel safe because safety is so important. We need psychological safety for anything we do even to have a conversation. We need psychological safety. So imagine having a >> I >> systems run our lives without having that psychological safety. That's bad news for all of us, right? And so we really need to focus on the trust. And we need to focus on our ability to trust the data or a right to help us trust the data or surface the issues that are causing the trust. >> Janet, what a pleasure to have you back on the Cube. I wish we had more time to keep talking, but it's I can't wait till we talk to you next year because what you guys are doing and also your pact, true passion for data science for trust and a I for good is palpable. So thank you so much for carving out some time to stop by the program. Thank you. It's my pleasure. We want to thank you for watching the Cuba and Lisa Martin live at Stanford for the fourth annual Women in Data Science conference. We back after a short break.

>> Announcer: From Milpitas, California at the edge of Silicon Valley, it's The CUBE. Covering autonomous vehicles. Brought to you by Western Digital. >> Hey, welcome back everybody. Jeff Frick here with The Cube. We are at the Autotech Council Autonomous Vehicle event at Western Digital. Part of our Data Makes Possible Program with Western Digital where we're looking at all these cool applications and a lot of cutting edge technology that at the end of the day, it's data dependent and data's got to sit somewhere. But really what's interesting here is that the data, and more and more of the data is moving out to the edge and edge computing and nowhere is that more apparent than in autonomous vesicles so we're really excited to have maybe the best title at Western Digital, I don't know. Chris Bergey, VP of Product Marketing. That's not so special, but all the areas that he's involved with: mobile, compute, automotive, connected homes, smart cities, and if that wasn't enough, industrial IOT. Chris, you must be a busy guy. >> Hey, we're having a lot of fun here. This data world is an exciting place to be right now. >> so we're her at the Autonomous Vehicle event. We could talk about smart cities, which is pretty interesting, actually ties to it and internet of things and industrial internets, but what are some of the really unique challenges in autonomous vehicles that most people probably aren't thinking of? >> Well, I think that we all understand that really, autonomous vehicles are being made possible by just the immense amount of sensors that are being put into the car. Not much different than as our smartphones or our phones evolved from really not having a lot of sensors to today's smartphones have many, many sensors. Whether it's sensing your face, gyroscopes, GPS, all these kind of things. The car is having the exact thing happen but many, many more sensors. And, of course, those sensors just drive a tremendous amount of data and then it's really about trying to pull the intelligence out of that data and that's really what the whole artificial intelligence or autonomous is really trying to do is, okay, we've got all this data, how do I understand what's happening in the autonomous vehicle in a very short period of time? >> Right, and there's two really big factors that you've talked about and some of the other things that you've done. I did some homework and one of them is the metadata around the data, so there's the raw data itself that's coming off those sensors, but the metadata is a whole nother level, and a big level, and even more importantly is the context. What is the context of that data and without context, it's just data. It's not really intelligence or smarts or things you can do anything about so that baseline sensor data gets amplified significantly in terms of actually doing anything with that information. >> That's correct. I think one of the examples I give that's easier for people to understand is surveillance, right? We're very familiar with walking into a retail store where there's surveillance cameras and they're recording in the case that maybe there's a theft or something goes wrong, but there's so much data there that's not acutely being processed, right? How may people walked into the store? What was the average time a person came to the store? How many men? How many women? That's the context of the data and that's what's really would be very valuable if you were, say, an owner of the store or a regional manager. So that's really pulling the context out of the raw data. And in the car example, autonomous vehicles, hey, there's going to be something, my sensors are seeing something, and then, of course, you'd use multiple sensors. That's the sensor fusion between them of, "Hey, that's a person, that's a deer, oh, don't worry, "that's a car moving alongside of us and he's "staying in his lane." Those are the types of decisions we're making with this data and that's the context. >> Right, and even they had in the earlier presentation today the reflection of the car off the side of a bus, I mean, these are the nuance things that aren't necessarily obvious when you first start exploring. >> And we're dealing with human life, I mean, so obviously it needs to be right 99.999 plus percent. So that's the challenge, right? It's the corner cases and I think that's what we see with autonomous vehicles. It's really exciting to see the developments going on and, of course, there's been a couple challenges, but we just have so much learning to do to really get to that fifth nine or whatever it is from a probability point of view. And that's where we'll continue to work on those corner cases, but the technology is coming along so fast, it's just mind-boggling how quickly we are starting to attack these more difficult challenges. And we'll get there but it's going to take time like anything. >> The other really important thing, especially now where we're in the rise of Cloud, if you will. Amazon is going bananas. Google Cloud Platform, Microsoft Azure, so we're seeing this huge move of Cloud and enterprise IT. But in a car, right, there's this little thing called latency and this other thing called physics where you've got a real issue when you have to make a quick decision based on data and those sensors when something jumps out in front of the car. So really, the rise of edge computing and moving so much of that stored compute and intelligence into the vehicle and then deciding what goes back to the car to retrain the algorithm. So it's really a shift to back out to the edge, if you will, dependent because of this latency issue. >> Yeah, I mean, they're very complimentary, right? But there's a lot of decisions you can make locally and, obviously, there's a lot of advantages in doing that. Latency being one of them, but just cost of communications and again, what people don't necessarily understand is how big this data is. You see statistics thrown out there, one gigabit per second, two gigabits per second. I mean, that is just massive data. At the end of the day, actually, in some of the development, it's pretty interesting that we have the car developers actually FedExing the terabyte drives that they've captured data because it's the easiest way for them to actually transfer the data. I mean, people think, "Oh, internet connectivity, no problem." You try to ship 80 terabytes in a cost effective manner, FedEx ends up being the best shot right now. So it's pretty interesting. >> The old sneaker, that is pretty funny. But the quantities of this data are so big. I was teasing you on Twitter earlier today. I think we took it up to an xobyte, a zedobyte, a yodabyte, and then the crowd responded. No, it's a brontosaurousbyte is even bigger than a yodabyte. We were at Flink Forward earlier this week and really this whole idea of stream processing, it's really taking new approaches to data processing. You'll be able to take all that stuff in in real time, which probably state of the market now is financial trading and advertising markets. But to do that now in a car where if you make a mistake, there's really significant consequences. It's a really different challenge. >> It is and again, that's really this advent of the sensor data, right? The sensor data is going to swamp probably every other data set that's in the world, but a lot of it's not interesting because you don't know when that interesting event is going to happen. So what you actually find is that you try to put it's intelligence as close as you can to the data, end storage, and again, storage may be 30 seconds to if you had an accident, you want to be able to go back 30 seconds. It may be lifetimes. So just thinking about these data flows and what's the half life of the data relative to the value? But what we're actually finding with many of the machine learning is that data we thought was not valuable, data we thought, "Oh, we have the right amount of granularity," now with machine learning we're going back and saying, "Oh, why didn't we record at an even higher granularity?" We could have pulled out more of these trends or more of these corner cases. So I think that's one of the challenges enterprise are going through right now is that everyone's so scared of getting rid of any data, yet there's just tremendous data growth. And we're sitting right here in the middle of it at Western Digital. >> Well, thankfully for you guys, you're going to store all that data and it is really important, though, because it used to be, it's funny to me. It used to be a sample of things that happened in the past is how you would make your decisions. Now it's not a sample, it's all of what's happening now and hopefully you can make a decision while you still have time to have an impact. So it's a very different world but sampling is going away when, in theory, you don't know what you're going to need that data for and you have the ability to store it. >> Making real-time decisions but then also learning how to use that decision to make better decisions in the future. That's really where Silicon Valley's focused right now. >> All right, Chris, well you're a busy guy so we're going to let you get back to it because you also have to do IOT and industrial internet and mobile an compute. So thanks for taking ... >> And I try to eat in between there too. >> And you try to eat and hopefully see your kids Friday night, so hopefully you'll take >> Absolutely. your wife out to a movie tonight. >> All right, Chris, great to see you. Thanks for taking a few minutes. >> Chris: Thank you very much. >> All right, I'm Jeff Frick. You're watching The CUBE from Autotech Council Autonomous Vehicle event. Thanks for watching.

>> Announcer: From Milpitas, California at the edge of Silicon Valley, it's theCUBE, covering autonomous vehicles. Brought to you by Western Digital. >> Hey, welcome back everybody. Jeff Frick, here with theCUBE. We're in Milpitas, California, at Western Digital, at the Autotech Council Autonomous Vehicle Event. About 300 people, really deep into this space. It's a developing ecosystem. You know, we think about Tesla, that's kind of got a complete, closed system. But there's a whole ecosystem of other companies getting into the autonomous vehicle space, and as was mentioned in the keynote, there are, literally thousands of problems. A great opportunity for startups. So we're excited to have Oded Sagee, he's a senior director of product marketing from Western Digital. Oded, great to see you. >> Thank you very much, Jeff. >> So you were just on the panel and, really that was a big topic, is there are thousands of problems to solve and this ecosystem's trying to come together, but it's complicated, right? It's not just the big car manufacturers anymore, and the tier one providers, but there's this whole ecosystem that's now growing up to try to solve these problems. So what are you seeing from your point of view? >> Yes, correct. So, definitely in the past automotive was a tough market to play in, but it was simple from the amount of players and people you needed to talk to to design your product inside. With the disruption of connectivity, smart vehicles, even before autonomous, there are so many new systems in the car now that generate data or consume data. And so, for us, to kind of figure out what's the use case, right? How is this going to look in the future? Who's going to define it? Who's going to buy it? Who's going to pay for it? It has become more and more complex. Happily, storage is in the center of all this. >> Jeff: Right. >> So we get a seat at the table and everyone wants to talk to us, but yes, it's a very big ecosystem now. And trying to resolve that problem, it's going to take some time. >> So what are some of the unique characteristics, from a storage point of view, that you have to worry about? Obviously environmental jumps out. We had the guy on before talking about bumpy roads, you know, the huge impacts on vibration. And now you spent a lot of money for a Toughbook back in the day to put a laptop in a cop car, this is a whole other level of expense, investment, and data flow. >> Right. So, for us, I think with all this disruption happening of full autonomous, people are, very much focused on making that autonomous work, right? So, for them it's all about connectivity, it's all about the sensor, whether it's Lidar, or, you know, cameras. Just making that work, right? All the algorithms and the software. And so, for them storage, currently is an afterthought, right? They were saying, once we meet mass production we'll just go and buy some storage and everything's going to be fine. So while they're prototyping, right? They can use any storage that they want. But, if you think about a full autonomous vehicle out there driving, not two hours a day like we are driving today, right? 20 hours a day, from cold to hot, going through areas without connectivity. Suddenly, the storage requirements are very, very different. And this is what we're trying to drive and explain that, if we don't design the future storage solutions today, What's going to end up, is that people are going to pay much more for storage just to make a basic use case work. >> Right. >> But if we start working now, and I'm talking about five, seven years out, we can have affordable solutions to make those business models work. >> And is that resonating in the industry, or are they just too focused on, you know, better cameras? >> It definitely does, but as companies change, right? So let's just take the car makers for a second. They didn't necessarily have a CTO in place, right? To drive engineering and semi-conductor. So you got to find those figures, and you got to start working and educating them. It definitely resonates if you have the right person. Once you find him, yes, it's on the list of priority. So we need to push. But it is happening. Yes, it is resonating. >> And it's so different because you do have this edge case. You have so much data being collected out in the field, if you will, within that vehicle. Some, to go back to the cloud, but you've got latency is always an issue, right? For safety. So, a little different storage challenge. So are there significant design thoughts that you guys are bringing into play on why this is so different and what is it going to take to really have kind of an optimal solution for autonomous vehicles? >> Yes, definitely there are a couple of vectors I would say, or knobs we need to work on. One of them is temperature. So, again vehicles do tend to go between hot and cold. Unlike many other components that just need to make sure that they operate between hot and cold, we actually have a big challenge on keeping data being accurate between hot and cold. So if you program cold and read hot and vice versa, data gets corrupted. >> Oh, even within the structures within the media? >> Yes. >> Okay. >> And people don't know that. So, for us to figure out, what's the temperature range that the car, through its lifetime, is going to go through. And make sure that we meet the use case, that's a big one. What we call the endurance on the cycling of the storage, again, if you cannot rely on connectivity, cannot rely on cloud because of latency, you need to record a lot of data in the car. So, again, a car drives for seven years, 15 years, and you want to record constantly, how much do you need to record? We don't necessarily have the technology today to meet that use case and we need to work with the ecosystem, in figuring it out. So these are just two examples. >> And I would imagine clean power, as you're saying these things, but they can need others. You're not in daddy's data center anymore. This is a pretty harsh environment, I would imagine. >> Very harsh. >> Ugly power, inconsistent power, turning off the car before everything is spun down. There's all kinds of little, kind of environmental impacts in that whole realm that you would never think of in, kind of a typical data center, for instance. >> Correct. And even, you touched power, that's very interesting because even some people think, oh, there's not power limitation in a car. You can just enjoy how much power you want. Actually, it's very, very sensitive. The battery, if you think about an EV car now has so many components to run and so even the power consumption, right? Just the energy that you need to consume is becoming critical for each, and every component >> in the vehicle. >> Right. And it's everybody's AI comparison, right? Is if Kasparov had to fight the computer with the same amount of power, it wouldn't have been much of a match. So the power to run all this AI stuff is not insignificant, so it is going to be a huge drain on these electric vehicles. Pretty exciting times. So when you get up in the morning, what's the biggest thing, when you talk to people about autonomous vehicles, that they just don't get? That people should really be thinking about. >> Yeah, so it goes back to some of the things we've discussed. Definitely, again, we're seeing the use cases change. We are working again with the broad ecosystem to explain the fundamental challenges that we have, right? What is our design cycle? What are the challenges that we have? So we start with educating the ecosystem, so they know what we have. And from that we trigger a discussion because they realize, oh, okay, because I do have a use case that, probably, you don't have a solution for, how do we go together? And we're doing it across the board. It's not only happening in automotive. It's happening in surveillance. It's happening in the home space. A lot of people don't know, but the home space, if you think about it, again, set-top boxes used to be huge, sat outside in the room. People are moving to these sticks, right? And they're behind the TV and they have no ventilation and they're small and they record all the time. And they get to temperatures that we've never seen in the past. So we even need to educate the telcos of the world, the set-top box makers. Everything is changing. Automotive is definitely ahead in a lot of innovation and disruption, but everything is changing for us. >> Right, a lot of those are fond of just the bright shiny object that everybody can see, right? We can't necessarily see a lot of IOT that GE's putting in to connect their factories. Alright, Oded, well thanks for taking a few minutes out of your busy day and I really appreciate the insight. >> Thank you very much. >> All right, he's Oded, I'm Jeff, You're watching theCUBE from Western Digital at The Autonomous Vehicle Event for the Autotech Council. Thanks for watching. Catch you next time. (electronic music)

(inspirational music) >> Hey, welcome back everybody. Jeff Frick here with theCUBE. We are in our Palo Alto studio. The conference season hasn't really kicked off yet into full swing so we can do a lot more kind of intimate stuff here in the studio, for a CUBE Conversation. And we're really excited to have a many time CUBE alum on, and a new guest, both from Western Digital. So Dave Tang, Senior Vice President at Western Digital. Great to see you again, Dave. >> Great to be here, Jeff. >> Absolutely and Martin Fink, he is the Chief Technology Officer at Western Digital, a longtime HP alum. I'm sure people recognized you from that and our great machine keynotes we were talking about it. So great to finally meet you, Martin. >> Thank you, nice to be here. >> Absolutely, so you guys are here talking about and we've got an ongoing program actually with Western Digital about Data Makes Possible, right. With all the things that are going on in tech at the end of the day, right, there's data, it's got to be stored somewhere and then of course there's processes and things going on. We've been exploring media and entertainment, sports, healthcare, autonomous vehicles, you know. All the places that this continues to reach out and it's such a fun project because you guys are a rising tide, lifts all boats, kind of company and really enjoy watching this whole ecosystem grow. So I really want to thank you for that. But now there's some new things that we want to talk about that you guys are doing to continue really in that same theme, and that's the support of this RISC-V. So first off, for people who have no idea, what is RISC-V? Let's jump into that and then kind of what is the announcement and why it's important. >> Sure, so RISC-V is an, you know, the tagline is, it's an open source instruction set architecture. So what does that mean, just so people can kind of understand. So today the world is dominated by two instruction set architectures. For the most part the, we'll call the desktop enterprise world is dominated by the Intel instruction set architecture and that's what's in most PCs, what people talk about as x86. And then the embedded and mobile space tends to be dominated by Arm, or by Arm Holdings. And so both of those are great architectures but they're also proprietary, they're owned by their respective companies. So RISC-V is essentially a third entrant, we'll say, into this world, but the distinction is that it's completely open source. So everything about the instruction set is available to all and anybody can implement it. We can all share the implementations. We can share the code that makes up that instruction set architecture, and very importantly for us and part of our motivation is the freedom to innovate. So we now have the ability to modify the instruction set or change the implementation of the instruction set, to optimize it for our devices and our storage and our drives, etc. >> So is this the first kind of open source play in microprocessor architecture? >> No, there's been other attempts at this. OpenSpark kind of comes to mind, and things like that, but the ability to get a community of individuals to kind of rally around this in a meaningful way has really been a challenge. And so I'd say that right now, RISC-V presents probably the best sort of clean slate, let's take some thing new to the market out there. >> So open source, obviously we've seen you know, take over the software world, first in the operating system which everybody is familiar with Linux but then we see it time and time again in different applications, Hadoop. I mean, there's just a proliferation of open source projects. The benefits are tremendous. Pretty easy to ascertain at a typical software case, how is that going to be applied do you think within the microprocessor world? >> So it's a little bit different. When we're talking about open source hardware or open source chips and microprocessors, you're dealing with a physical device. So even though you can open source all of the designs and the code associated with that device, you still have to fabricate it. You still have to create a physical design and you still have to call up a fab and say, will you make this for me at these particular volumes? And so that's the difference. So there are some differences between open source software where it's, you know, you create the bits and then you distribute those bits through the Internet and all is good. Whereas here, you still have a physical need to fabricate something. >> Now, how much more flexibility can you do then for the output when you can actually impact the architecture as opposed to just creating a custom chip design, on top of somebody else's architecture? >> Well, let me give you probably a really simple, concrete example that kind of people can internalize of some of our motivation behind this, because that might sort of help get people through this. If you think of a very typical surveillance application, you have a camera pointed into a room or a hallway. The reality is we're basically grabbing a ton of video frames but very few of them change, right? So the typical surveillance application is it never changes and you really want, only know when stuff changes. Well, today, in very simple terms, all of those frames get routed up to some big server somewhere and that server spends a lot of time trying to figure out, okay have I got a frame that changed? Have I got a frame that changed, and so on. And then eventually it'll find maybe two or three or five frames that have got something interesting. So in the world what we're trying to do is to say, okay well why don't we take that, find no changes, and push that right down to the device? So we basically store all those frames, why don't we go figure out all the frames that mean nothing, and only ship up to that big bad server the frames that have something interesting and something you want to go analyze and do some work on? So that's a very typical application that's quite meaningful because we can do all of that work at the device. We can eliminate shipping a whole bunch of data to where it's just going to get discarded anyways, and we can allow the end customer to really focus on the data that matters, and get some intelligence. >> And that's critical as we get more and more immersed in a data-centric world, where we have realtime applications like Martin described as well as large data-centric applications like of course, big data analytics, but also training for AI systems or machine learning. These workloads are going to become more and more diverse and they're going to need more specialized architectures and more specialized processing. So big data is getting bigger and faster and these realtime fast data applications are getting faster and bigger. So we need ways to contend with that, that really go beyond what's available with general purpose architectures. >> So that's a great point because if we take this example of video frames, now if I can build a processor that is customized to only do that, that's the only thing it does. It can be very low power, very efficient, and do that one thing very very well, and the cost adder, if you want to call it that, to the device where we put it, is a tiny fraction, but the cost savings of the overall solution is significant. So this ability to customize the instruction set to only do what you need it to do for that very special purpose, that's gold. >> So I just wanted to, Dave, we've talked about a lot of interesting innovations that you guys have come up with over the years, with the helium launch. Which I don't know, a couple, two, three years ago, you were just at the MAMR event, really energy assisted recording. So this is really kind of foundational within the storage and the media itself and how you guys do better and take advantage of evolving land space. This is a kind of a different play for Western Digital, this isn't a direct kind of improvement in the way that storage media and architecture works but this is really more of, I'm going to ask you. What is the Western Digital play here? Why is this an important space for you guys in your core storage business? >> Well we're really broadening our focus to really develop and innovate around technologies that really help the world extract more value from data as a whole, right. So it's way beyond storage these days, right. We're looking for better ways to capture, preserve, access, and transform the data. And unless you transform it, you can't really extract the value out of it so as we see all these new applications for data and the vast possibilities for data, we really want to pave the path and help the industry innovate to bring all those applications to reality. >> It's interesting too because one of the great topics always in computing is you know, you got compute and store, which has to go to which, right. And nobody wants to move a lot of data, that's hard and may or may not be easy to get compute. Especially these IoT applications, remote devices, tough conditions and power, which we mentioned a little bit before we went on air. So the the landscape for the for the need for compute and store in networking is radically changing than either the desktop or what we're seeing a consolidation in clouds. So what's interesting here, where does the scale come, right? At the end of the day, scale always wins. And that's where we've seen historically where the general-purpose microprocessor architectures is dominated but used to be a slew of specialty purpose architectures but now there's an opportunity to bring scale to this. So how does that scale game continue to evolve? >> So it's a great point that scale does matter and we've seen that repeatedly and so it's a significant part of the reason why we decided to go early with a significant commitment was to tell the world that we were bringing scale to the equation. And so what we communicated to the marketplace is we ship on the order of a billion processor cores a year, most people don't realize that all of our devices from USB sticks to hard drives, all have processors on them. And so we said, hey we're going to basically go all-in and go big and that translates into a billion cores that we ship every year and we're going to go on a program to essentially migrate all of those cores to RISC-V. It'll take a few years to get there but we'll migrate all of those cores and so we basically were signaling to the market, hey scale is now here. Scale is here, you can make the investments, you can go forward, you can make that commitment to RISC-V because essentially we've got your back. >> So just to make sure we get that clear. So you guys have announced that you're going to slowly migrate over time your micro processors that power your devices to the tune of approximately a billion with a B, cores per year to this new architecture. >> That is correct. >> And has that started? >> So the design has started. So we have started to design and develop our first two cores but the actual manifestation into devices probably in the early stage of 2020. >> Okay, okay. But that's a pretty significant commitment and again, the ideas you explicitly said it's a signal to the ecosystem, this is worth your investment because there is some scale here. >> Martin: That's right. >> Yeah, pretty exciting. And how do you think it's going to open up the ability for you to do new things with your devices that you before either couldn't do or we're too expensive with dollars or power. >> Martin: So we're going to step and iterate through this and one key point here is a lot of people tend to want to start in this processor world at the very high end, right. I'm going to go take on a Xeon processor or something like that. It's not what we're doing. We're basically saying, we're going to go at the small end, the tiny end where power matters. Power matters a lot in our devices and where can we achieve the optimum combination of power and performance. So even in our small devices like a USB stick or a client SSD or something like that, if we can reduce power consumption and even just maintain performance that's a huge win for our customers, you know. If you think about your laptop and if I reduce the power consumption of that SSD in there so that you have longer battery life and you can get you know through the day better, that's a huge win, right. And I don't impact performance in the process, that's a huge win. So what we do, what we're doing right now is we're developing the cores based on the RISC-V architecture and then what we're going to do is once we've got that sort of design, sort of complete is we want to take all of the typical client workloads and profile them on that. Then we want to find out, okay where are the hot spots? What are the two or three things that are really consuming all the power and how do we go optimize, by either creating two or three instructions or by optimizing the micro architecture for an existing instruction. And then iterate through that a few times so that we really get a big win, even at the very low end of the spectrum and then we just iterate through that with time. >> We're in a unique position I think in that the technologies that we develop span everything from the actual media where the bits are stored, whether it's solid-state flash or rotating magnetic disk and the recording heads. We take those technologies and build them all the way up into devices and platforms and full-fledged data center systems. And if we can optimize and tune all the way from that core media level all the way up through into the system level, we can deliver significantly higher value, we believe, to the marketplace. So this is the start of that, that enables us to customize command sets and optimize the flow of data so that we can we can allow users to access it when and where they need it. >> So I think there's another actually really cool point, which goes back to the open source nature of this and we try to be very clear about this. We're not going to develop our cores for all applications. We want the world to develop all sorts of different cores. And so for many applications somebody else might come in and say, hey we've got a really cool core. So one of the companies we've partnered with and invested in for example, is Esperanto. They've actually decided to go at the high end and do a machine learning accelerator. Hey, maybe we'll use that for some machine learning applications in our system level performance. So we don't have to do it all but we've got a common architecture across the portfolio and that speaks to that sort of open source nature of the RISC-V architecture is we want the world to get going. We want our competitors to get on board, we want partners, we want software providers, we want everybody on board. >> It's such a different ecosystem with open-source and the way the contributions are made and the way contributions are valued and the way that people can go find niches that are underserved. It's this really interesting kind of bifurcation of the market really, you don't really want to be in the big general-purpose middle anymore. That's not a great place to be, there's all kinds of specialty places where you can build the competence and with software and you know with, thank goodness for Moore's law decreasing prices of the power of the compute and now the cloud, which is basically always available. Really a exciting time to develop a myriad of different applications. >> Right and you talked before about scale in terms of points of implementation that will drive adoption and drive this to critical mass but there's another aspect of scale relative to the architecture within a single system that's also important that I think RISC-V helps to break down some barriers. Because with general purpose computer architectures, they assume a certain ratio of memory and storage and processing and bandwidth for interconnect and if you exceed those ratios, you have to add a whole new processor. Even though you don't need to need the processing capability, you need it for scale. So that's another great benefit of these new architectures is that the diversity of data needs where some are going to be large data sets, some are going to be small data sets that need need high bandwidth. You can customize and blend that recipe as you need to, you're not at the mercy of these fixed ratios. >> Yeah and I think you know it's so much of kind of what is cloud computing. And the atomic nature of it, that you can apply the ratios, the amount that you need as you need, you can change it on the fly, you can tone it up, tone it down. And I think the other interesting thing that you touched on is some of these new, which are now relatively special-purpose but are going to be general-purpose very soon in terms of machine learning and AI and applying those to different places and applying them closer to the problem. It's a very very interesting evolution of the landscape but what I want to do is kind of close on you Martin, especially because again kind of back to the machine. Not the machine specifically but you have been in the business of looking way down the road for a long time. So you came out, I'd looked at your LinkedIn, you retired for three months, congratulations. (laughs) Hope you got some my golf in but you came back to Western Digital so why did you come back? And as you look down the road a ways, what do you see that it excites you, that got you off that three-month little tour around the golf course and I'm sorry I had to tease about that. But what do you see? What are you excited about that you came back and got involved in an open source microprocessor project? >> So the the short answer was that, I saw the opportunity at Western Digital to be where data lives. So I had spent my entire career, will call it at the compute or the server side of things and the interesting thing is I had a very close relationship with SanDisk, which was acquired by Western Digital. And so I had, we'll call it an insider view, of what was possible there and so what triggered was essentially what we're talking about here was given that about half the world's data lands on Western Digital devices, taking that from a real position of strength in the marketplace and say, what could we go do to make data more intelligent and rather than start kind of at that server end and so that I saw that potential there and it was just incredible, so that's that's what made me want to join. >> Exciting times. Dave good get. (laughs) >> We're delighted to have Martin with us. >> All right, well we look forward to watch it evolve. We've got another another whole set of events we're going to do again together with Western Digital that we're excited about. Again, covering Data Makes Possible but you know kind of uplifting into the application space as a lot of the cool things that people are doing in innovation. So Martin, great to finally meet you and thanks for stopping by. >> Thanks for the time. >> David as always and I think we'll see in a month or so. >> Right, always a pleasure Jeff, thanks. >> All right Martin Fink, Dave Tang. I'm Jeff Frick, you're watching theCUBE. Thanks for watching, we'll catch you next time. (inspirational music)

>> Announcer: Live from San Jose, California, it's theCUBE covering innovating to fuel the next decade of big data. Brought to you by Western Digital. >> Hey, welcome back everybody. Jeff Frick, here with theCUBE. We are at the Western Digital Headquarters in San Jose, California. It's the Al-Mady Campus. A historic campus. It's had a lot of great innovation, especially in hard drives for years and years and years. This event's called Innovating to Fuel the Next Data Big Data. And we're excited to have a big brain on. We like to get smart people who's been watching this story for a while and will give us a little bit of historical perspective. It's John Rydning. He is the Research Vice President for Hard Drives for IEC. John, Welcome. >> Thank you, Jeff. >> Absolutely. So, what is your take on today's announcement? >> I think it's our very meaningful announcement, especially when you consider that the previous BIGIT Technology announcement for the industry was Helium, about four or five years ago. But, really, the last big technology announcement prior to that was back in 2005, 2006, when the industry announced making this transition to what they called at that time, "Perpendicular Magnetic Recording." And when that was announced it was kind of a similar problem at that time in the industry that we have today, where the industry was just having a difficult time putting more data on each disc inside that drive. And, so they kind of hit this technology wall. And they announced Perpendicular Magnetic Recording and it really put them on a new S curve in terms of their ability to pack more data on each disc and just kind of put it in some perspective. So, after they announce Perpendicular Magnetic Recording, the capacity per disc increased about 30% a year for about five years. And then over, really, a ten year period, increased about an average of about 20% a year. And, so today's announcement is I see a lot of parallels to that. You know, back when Perpendicular Magnetic Recording was announced, really they build. They increased the capacity per platter was growing very slowly. That's where we are today. And with this announcement of MAMR Technology the direction that Western Digital's choosing really could put the industry on a new S curve and putting in terms of putting more capacity, storage capacity on each one of those discs. >> It's interesting. Always reminds me kind of back to the OS in Microsoft in Intel battles. Right? Intel would come out with a new chip and then Microsoft would make a bigger OS and they go back and back and forth and back and forth. >> John: Yeah, that's very >> And we're seeing that here, right? Cuz the demands for the data are growing exponentially. I think one of the numbers that was thrown out earlier today that the data thrown off by people and the data thrown off by machines is so exponentially larger than the data thrown off by business, which has been kind of the big driver of IT spin. And it's really changing. >> It's a huge fundamental shift. It really is >> They had to do something. Right? >> Yeah, the demand for a storage capacity by these large data centers is just phenomenal and yet at the same time, they don't want to just keep building new data center buildings. And putting more and more racks. They want to put more storage density in that footprint inside that building. So, that's what's really pushing the demand for these higher capacity storage devices. They want to really increase the storage capacity per cubic meter. >> Right, right. >> Inside these data centers. >> It's also just fascinating that our expectation is that they're going to somehow pull it off, right? Our expectation that Moore's laws continue, things are going to get better, faster, cheaper, and bigger. But, back in the back room, somebody's actually got to figure out how to do it. And as you said, we hit these kind of seminal moments where >> Yeah, that's right. >> You do get on a new S curve, and without that it does flatten out over time. >> You know, what's interesting though, Jeff, is really about the time that Perpendicular Magnetic Recording was announced way back in 2005, 2006, the industry was really, already at that time, talking about these thermal assist technologies like MAMR that Western Digital announced today. And it's always been a little bit of a question for those folks that are either in the industry or watching the industry, like IDC. And maybe even even more importantly for some of the HDD industry customers. They're kind of wondering, so what's really going to be the next technology race horse that takes us to that next capacity point? And it's always been a bit of a horse race between HAMR and MAMR. And there's been this lack of clarity or kind of a huge question mark hanging over the industry about which one is it going to be. And Western Digital certainly put a stake in the ground today that they see MAMR as that next technology for the future. >> (mumbles words) Just read a quote today (rushes through name) key alumni just took a new job. And he's got a pin tweet at the top of his thing. And he says, "The smart man looks for ways "To solve the problem. "Or looks at new solutions. "The wise man really spends his time studying the problem." >> I like that. >> And it's really interesting here cuz it seems kind of obvious there. Heat's never necessarily a good thing with electronics and data centers as you mentioned trying to get efficiency up. There's pressure as these things have become huge, energy consumption machines. That said, they're relatively efficient, based on other means that we've been doing they compute and the demand for this compute continues to increase, increase, increase, increase. >> Absolutely >> So, as you kind of look forward, is there anything kind of? Any gems in the numbers that maybe those of us at a layman level are kind of a first read are missing that we should really be paying attention that give us a little bit of a clue of what the feature looks like? >> Well, there's a couple of major trends going on. One is that, at least for the hard drive industry, if you kind of look back the last ten years or so, a pretty significant percentage of the revenue that they've generated a pretty good percentage of the petabytes that they ship have really gone into the PC market. And that's fundamentally shifting. And, so now it's really the data centers, so that by the time you get to 2020, 2021, about 60 plus percent of the petabytes that the industry's shipping is going into data centers, where if you look back a few years ago, 60% was going into PCs. That's a big, big change for the industry. And it's really that kind of change that's pushing the need for these higher capacity hard drives. >> Jeff: Right. >> So, that's, I think, one of the biggest shifts has taking place. >> Well, the other thing that's interesting in that comment because we know scale drives innovation better than anything and clearly Intel microprocessors rode the PC boom to get out scale to drive the innovation. And, so if you're saying, now, that the biggest scale is happening in the data center Then, that's a tremendous force for innovation in there versus Flash, which is really piggy-backing on the growth of these jobs, because that's where it's getting it's scale. So, when you look at kind of the Flash hard drive comparison, right? Obviously, Flash is the shiny new toy getting a lot of buzz over the last couple years. Western Digital has a play across the portfolio, but the announcement earlier today said, you're still going to have like this TenX cost differentiation. >> Yeah, that's right. >> Even through, I think it was 20, 25. I don't want to say what the numbers were. Over a long period of time. You see that kind of continuing DC&E kind of conflict between those two? Or is there a pretty clear stratification between what's going to go into Flash systems, or what's going to hard drives? >> That's a great question, now. So, even in the very large HyperScale data centers and we definitely see where Flash and hard disk drives are very complimentary. They're really addressing different challenges, different problems, and so I think one of the charts that we saw today at the briefing really is something that we agree with strongly at IDC. Today, maybe, about 7% or 8% of all of the combined HDD SSD petabyte shipped for enterprise are SSD petabytes. And then, that grows to maybe ten. >> What was it? Like 7% you said? >> 6% to 7%. >> 6% to 7% okay. Yeah, so we still have 92, 93%, 94% of all petabytes that again are HDD SSD petabytes for enterprise. Those are still HDD petabytes. And even when you get out to 2020, 2021, again, still bought 90%. We agree with what Western Digital talked about today. About 90% of the combined HDD SSD petabytes that are shipping for enterprise continue to be HDD. So, we do see the two technologies very complementary. Talked about SSD is kind of getting their scale on PCs and that's true. They really are going to quickly continue to become a bigger slice of the storage devices attached to new PCs. But, in the data center you really need that bulk storage capacity, low cost capacity. And that's where we see that the two SSDs and HDDs are going to live together for a long time. >> Yeah, and as we said the conflict barrier, complimentary nature of the two different applications are very different. You need the big data to build the models, to run the algorithms, to do stuff. But, at the same time, you need the fast data that's coming in. You need the real time analytics to make modifications to the algorithms and learn from the algorithms >> That's right, yeah. It's the two of those things together that are one plus one makes three type of solution. Exactly, and especially to address latency. Everybody wants their data fast. When you type something into Google, you want your response right away. And that's where SSDs really come into play, but when you do deep searches, you're looking through a lot of data that has been collected over years and a lot of that's probably sitting on hard disc drives. >> Yeah. The last piece of the puzzle, I just want to you to address before we sign off, That was an interesting point is that not just necessarily the technology story, but the ecosystem story. And I thought that was really kind of, I thought, the most interesting part of the MAMR announcement was that it fits in the same form factor, there's no change to OS, there's no kind of change in the ecosystem components in which you plug this in. >> Yeah, that's right. It's just you take out the smaller drive, the 10, or the 12, or whatever, or 14 I guess is coming up. And plug in. They showed a picture of a 40 terabyte drive. >> Right. >> You know, that's the other part of the story that maybe doesn't get as much play as it should. You're playing in an ecosystem. You can't just come up with this completely, kind of independent, radical, new thing, unless it'S so radical that people are willing to swap out their existing infrastructure. >> I completely agree. It's can be very difficult for the customer to figure out how to adopt some of these new technologies and actually, the hard disk drive industry has thrown a couple of technologies at their customers over the past five, six years, that have been a little challenging for them to adopt. So, one was when the industry went from a native 512 by sectors to 4K sectors. Seems like a pretty small change that you're making inside the drive, but it actually presented some big challenges for some of the enterprise customers. And even the single magnetic recording technologies. So, it has a way to get more data on the disc, and Western Digital certainly talked about that today. But, for the customer trying to plug and play that into a system and SMR technology actually created some real challenges for them to figure out how to adopt that. So, I agree that what was shown today about the MAMR technology is definitely a plug and play. >> Alright, we'll give you the last word as people are driving away today from the headquarters. They got a bumper sticker as to why this is so important. What's it say on the bumper sticker about MAMR? It says that we continue to get more capacity at a lower cost. >> (chuckles) Isn't that just always the goal? >> I agree. >> (chuckles) Alright, well thank you for stopping by and sharing your insight. Really appreciate it. >> Thanks, Jeff. >> Alright. Jeff Frick here at Western Digital. You're watching theCUBE! Thanks for watching. (futuristic beat)

>> Announcer: Live from San Jose, California, it's theCUBE, covering Innovating to Fuel the Next Decade of Big Data, brought to you by Western Digital. >> Hey welcome back everybody, Jeff Frick here with theCUBE. We're at Western Digital at their global headquarters in San Jose, California, it's the Almaden campus. This campus has a long history of innovation, and we're excited to be here, and probably have the smartest person in the building, if not the county, area code and zip code. I love to embarrass here, Janet George, she is the Fellow and Chief Data Scientist for Western Digital. We saw you at Women in Data Science, you were just at Grace Hopper, you're everywhere and get to get a chance to sit down again. >> Thank you Jeff, I appreciate it very much. >> So as a data scientist, today's announcement about MAMR, how does that make you feel, why is this exciting, how is this going to make you be more successful in your job and more importantly, the areas in which you study? >> So today's announcement is actually a breakthrough announcement, both in the field of machine learning and AI, because we've been on this data journey, and we have been very selectively storing data on our storage devices, and the selection is actually coming from the preconstructed queries that we do with business data, and now we no longer have to preconstruct these queries. We can store the data at scale in raw form. We don't even have to worry about the format or the schema of the data. We can look at the schema dynamically as the data grows within the storage and within the applications. >> Right, cause there's been two things, right. Before data was bad 'cause it was expensive to store >> Yes. >> Now suddenly we want to store it 'cause we know data is good, but even then, it still can be expensive, but you know, we've got this concept of data lakes and data swamps and data all kind of oceans, pick your favorite metaphor, but we want the data 'cause we're not really sure what we're going to do with it, and I think what's interesting that you said earlier today, is it was schema on write, then we evolved to schema on read, which was all the rage at Hadoop Summit a couple years ago, but you're talking about the whole next generation, which is an evolving dynamic schema >> Exactly. >> Based whatever happens to drive that query at the time. >> Exactly, exactly. So as we go through this journey, we are now getting independent of schema, we are decoupled from schema, and what we are finding out is we can capture data at its raw form, and we can do the learning at the raw form without human interference, in terms of transformation of the data and assigning a schema to that data. We got to understand the fidelity of the data, but we can train at scale from that data. So with massive amounts of training, the models already know to train itself from raw data. So now we are only talking about incremental learning, as the train model goes out into the field in production, and actually performs, now we are talking about how does the model learn, and this is where fast data plays a very big role. >> So that's interesting, 'cause you talked about that also earlier in your part of the presentation, kind of the fast data versus big data, which kind of maps the flash versus hard drive, and the two are not, it's not either or, but it's really both, because within the storage of the big data, you build the base foundations of the models, and then you can adapt, learn and grow, change with the fast data, with the streaming data on the front end, >> Exactly >> It's a whole new world. >> Exactly, so the fast data actually helps us after the training phase, right, and these are evolving architectures. This is part of your journey. As you come through the big data journey you experience this. But for fast data, what we are seeing is, these architectures like Lambda and Kappa are evolving, and especially the Lambda architecture is very interesting, because it allows for batch processing of historical data, and then it allows for what we call a high latency layer or a speed layer, where this data can then be promoted up the stack for serving purposes. And then Kappa architecture's where the data is being streamed near real time, bounded and unbounded streams of data. So this is again very important when we build machine learning and AI applications, because evolution is happening on the fly, learning is happening on the fly. Also, if you think about the learning, we are mimicking more and more on how humans learn. We don't really learn with very large chunks of data all at once, right? That's important for initially model training and model learning, but on a regular basis, we are learning with small chunks of data that are streamed to us near real time. >> Right, learning on the Delta. >> Learning on the Delta. >> So what is the bound versus the unbound? Unpack that a little bit. What does that mean? >> So what is bounded is basically saying, hey we are going to get certain amounts of data, so you're sizing the data for example. Unbounded is infinite streams of data coming to you. And so if your architecture can absorb infinite streams of data, like for example, the sensors constantly transmitting data to you, right? At that point you're not worried about whether you can store that data, you're simply worried about the fidelity of that data. But bounded would be saying, I'm going to send the data in chunks. You could also do bounded where you basically say, I'm going to pre-process the data a little bit just to see if the data's healthy, or if there is signal in the data. You don't want to find that out later as you're training, right? You're trying to figure that out up front. >> But it's funny, everything is ultimately bounded, it just depends on how you define the unit of time, right, 'cause you take it down to infinite zero, everything is frozen. But I love the example of the autonomous cars. We were at the event with, just talking about navigation just for autonomous cars. Goldman Sachs says it's going to be a seven billion dollar industry, and the great example that you used of the two systems working well together, 'cause is it the car centers or is it the map? >> Janet: That's right. >> And he says, well you know, you want to use the map, and the data from the map as much as you can to set the stage for the car driving down the road to give it some level of intelligence, but if today we happen to be paving lane number two on 101, and there's cones, now it's the real time data that's going to train the system. But the two have to work together, and the two are not autonomous and really can't work independent of each other. >> Yes. >> Pretty interesting. >> It makes perfect sense, right. And why it makes perfect sense is because first the autonomous cars have to learn to drive. Then the autonomous cars have to become an experienced driver. And the experience cannot be learned. It comes on the road. So one of the things I was watching was how insurance companies were doing testing on these cars, and they had a human, a human driving a car, and then an autonomous car. And the autonomous car, with the sensors, were predicting the behavior, every permutation and combination of how a bicycle would react to that car. It was almost predicting what the human on the bicycle would do, like jump in front of the car, and it got it right 80% of the cases. But a human driving a car, we're not sure how the bicycle is going to perform. We don't have peripheral vision, and we can't predict how the bicycle is going to perform, so we get it wrong. Now, we can't transmit that knowledge. If I'm a driver and I just encountered a bicycle, I can't transmit that knowledge to you. But a driverless car can learn, it can predict the behavior of the bicycle, and then it can transfer that information to a fleet of cars. So it's very powerful in where the learning can scale. >> Such a big part of the autonomous vehicle story that most people don't understand, that not only is the car driving down the road, but it's constantly measuring and modeling everything that's happening around it, including bikes, including pedestrians, including everything else, and whether it gets in a crash or not, it's still gathering that data and building the model and advancing the models, and I think that's, you know, people just don't talk about that enough. I want follow up on another topic. So we were both at Grace Hopper last week, which is a phenomenal experience, if you haven't been, go. Ill just leave it at that. But Dr. Fei-Fei Li gave one of the keynotes, and she made a really deep statement at the end of her keynote, and we were both talking about it before we turned the cameras on, which is, there's no question that AI is going to change the world, and it's changing the world today. The real question is, who are the people that are going to build the algorithms that train the AI? So you sit in your position here, with the power, both in the data and the tools and the compute that are available today, and this brand new world of AI and ML. How do you think about that? How does that make you feel about the opportunity to define the systems that drive the cars, et cetera. >> I think not just the diversity in data, but the diversity in the representation of that data are equally powerful. We need both. Because we cannot tackle diverse data, diverse experiences with only a single representation. We need multiple representation to be able to tackle that data. And this is how we will overcome bias of every sort. So it's not the question of who is going to build the AI models, it is a question of who is going to build the models, but not the question of will the AI models be built, because the AI models are already being built, but some of the models have biases into it from any kind of lack of representation. Like who's building the model, right? So I think it's very important. I think we have a powerful moment in history to change that, to make real impact. >> Because the trick is we all have bias. You can't do anything about it. We grew up in the world in which we grew up, we saw what we saw, we went to our schools, we had our family relationships et cetera. So everyone is locked into who they are. That's not the problem. The problem is the acceptance of bring in some other, (chuckles) and the combination will provide better outcomes, it's a proven scientific fact. >> I very much agree with that. I also think that having the freedom, having the choice to hear another person's conditioning, another person's experiences is very powerful, because that enriches our own experiences. Even if we are constrained, even if we are like that storage that has been structured and processed, we know that there's this other storage, and we can figure out how to get the freedom between the two point of views, right? And we have the freedom to choose. So that's very, very powerful, just having that freedom. >> So as we get ready to turn the calendar on 2017, which is hard to imagine it's true, it is. You look to 2018, what are some of your personal and professional priorities, what are you looking forward to, what are you working on, what's top of mind for Janet George? >> So right now I'm thinking about genetic algorithms, genetic machine learning algorithms. This has been around for a while, but I'll tell you where the power of genetic algorithms is, especially when you're creating powerful new technology memory cell. So when you start out trying to create a new technology memory cell, you have materials, material deformations, you have process, you have hundred permutation combination, and the genetic algorithms, we can quickly assign a cause function, and we can kill all the survival of the fittest, all that won't fit we can kill, arriving to the fastest, quickest new technology node, and then from there, we can scale that in mass production. So we can use these survival of the fittest mechanisms that evolution has used for a long period of time. So this is biology inspired. And using a cause function we can figure out how to get the best of every process, every technology, all the coupling effects, all the master effects of introducing a program voltage on a particular cell, reducing the program voltage on a particular cell, resetting and setting, and the neighboring effects, we can pull all that together, so 600, 700 permutation combination that we've been struggling on and not trying to figure out how to quickly narrow down to that perfect cell, which is the new technology node that we can then scale out into tens of millions of vehicles, right? >> Right, you're going to have to >> Getting to that spot. >> You're going to have to get me on the whiteboard on that one, Janet. That is amazing. Smart lady. >> Thank you. >> Thanks for taking a few minutes out of your time. Always great to catch up, and it was terrific to see you at Grace Hopper as well. >> Thank you, I really appreciate it, I appreciate it very much. >> All right, Janet George, I'm Jeff Frick. You are watching theCUBE. We're at Western Digital headquarters at Innovating to Fuel the Next Generation of Big Data. Thanks for watching.

>> Announcer: Live from San Jose, California, it's theCUBE, covering Innovating to Fuel the Next Decade of Big Data, brought to you by Western Digital. >> Hey welcome back everybody, Jeff Frick here with theCUBE. We're at Western Digital's headquarters in San Jose, California at the Almaden campus. Lot of innovation's been going on here, especially in storage for decades, and we're excited to be at this special press and analyst event that Western Digital put on today to announce some exciting new products. It's called Innovating to Fuel the Next Decade of Data. I'm super happy to have a long-time industry veteran, he just told me, 35 years, I don't know if I can tell (Mark laughs) that on air or not. He's Mark Grace, he's the Senior Vice President of Devices for Western Digital, Mar, great to have you on. >> Thanks Jeff, glad to be here. >> Absolutely, so you've seen this movie over and over and over, I mean that's one of the cool things about being in the Valley, is this relentless pace of innovation. So how does today's announcement stack up as you kind of look at this versus kind of where we've come from? >> Oh I think this is maybe one of the, as big as it comes, Jeff, to be honest. I think we've plotted a course now that I think was relatively uncertain for the hard drive industry and the data center, and plotted a course that I think we can speak clearly to the market, and clearly to customers about the value proposition for rotating magnetic storage for decades to come. >> Which is pretty interesting, 'cause, you know, rotating drives have been taking a hit over the last couple of years, right, flash has been kind of the sexy new kid on the block, so this is something new, >> Mark: It is. >> And a new S curve I think as John said. >> I agree, we're jumping onto a, we're extending the S curve, let's call it that. I think there's actually plenty of other S curve opportunities for us, but in this case, I think the industry, and I would say our customer base, we have been less than clear with those guys about how we see the future of rotating storage in the cloud and enterprise space, and I think today's announcement clarifies that and gives some confidence about architectural decisions relative to rotating storage going forward for a long time. >> Well I think it's pretty interesting, 'cause compared to the other technology that was highlighted, the other option, the HAMR versus the MAMR, this was a much more elegant, simpler way to add this new S curve into an existing ecosystem. >> You know, elegant's probably a good word for it, and it's always the best solution I would say. HAMR's been a push for many years. I can't remember the first time I heard about HAMR. It's still something we're going to continue to explore and invest in, but it has numerous hurdles compared to the simplicity and elegance, as you say, of MAMR, not the least of which is we're going to operate at normal ambient temperatures versus apply tremendous heat to try and energize the recording and the technologies. So any time you introduce extraordinary heat you face all kinds of ancillary engineering challenges, and this simplifies those challenges down to one critical innovation, which is the oscillator. >> Pretty interesting, 'cause it seems pretty obvious that heat's never a good thing. So it's curious that in the quest for this next S curve that the HAMR path was pursued for as long as it was, it sounds like, because it sounds like that's a pretty tough thing to overcome. >> Yeah, I think it initially presented perhaps the most longevity perhaps in early exploration days. I would say that HAMR has certainly received the most press as far as trying to assert it as the extending recording technology for enterprise HDDs. I would say we've invested for almost as long in MAMR, but we've been extremely quiet about it. This is kind of our nature. When we're ready to talk about something, you can kind of be sure we're ready to go with it, and ready to think about productization. So we're quite confident in what we're doing. >> But I'm curious from your perspective, having been in the business a long time, you know, we who are not directly building these magical machines, just now have come to expect that Moore's Law will contain, has zero to do with semiconductor physics anymore, it's really an attitude and this relentless pace of innovation that now is expected and taken for granted. You're on the other side, and have to face real physics and mechanical limitations of the media and the science and everything else. So is that something that gets you up every day >> Mark: Keeps me awake every night! >> Obviously keeps you awake at night and up every day. You've been doing it for 35 years, so there must be some appeal. >> Yeah. (laughs) >> But you know, it's a unique challenge, 'cause at the same time not only has it got to be better and faster and bigger, it's got to be cheaper, and it has been. So when you look at that, how does that kind of motivate you, the teams here, to deliver on that promise? >> Yeah, I mean in this case, we are a little bit defensive, in the sense of the flash expectations that you mentioned, and I think as we digest our news today, we'll be level setting a little bit more in a more balanced way the expectations for contribution from rotating magnetic storage and solid state storage to what I think is a more accurate picture of its future going forward in the enterprise and hyperscale space. To your point about just relentless innovation, a few of us were talking the other day in advance of this announcement that this MAMR adventure feels like the early days of PMR, perpendicular, the current recording technology. It feels like we understand a certain amount of distance ahead of us, and that's about this four-terabit per inch kind of distance, but it feels like the early days where we could only see so far but the road actually goes much further, and we're going to find more and more ways to extend this technology, and keep that order of magnitude cost advantage going from a hard drive standpoint versus flash. >> I wonder how this period compares to that, just to continue, in terms of on the demand side, 'cause you know, back in the day, the demand and the applications for these magical compute machines weren't near, I would presume, as pervasive as now, or am I missing the boat? 'Cause now clearly there is no shortage of demand for storage and compute. >> Yeah, depending on where you're coming from, you could take two different views of that. The engine that drove the scale of the hard drive industry to date has, a big piece of it in the long history of the hard drive industry has been the PC space. So you see that industry converting to flash and solid state storage more aggressively, and we embrace that, you know we're invested in flash and we have great products in that space, and we see that happening. The opportunity in the hyperscale and cloud space is we're only at the tip of the iceberg, and therefore I think, as we think about this generation, we think about it differently than those opportunities in terms of breadth of applications, PCs, and all that kind of create the foundation for the hard drive, but what we see here is the virtuous cycle of more storage, more economical storage begets more value proposition, more opportunities to integrate more data, more data collection, more storage. And that virtuous cycle seems to me that we're just getting started. So long live data, that's what we say. (both laugh) >> The other piece that I find interesting is before the PCs were the driver of scale relative to an enterprise data center, but with the hyperscale guys and the proliferation of cloud and actually the growth of PCs is slowing down dramatically, that it's kind of flipped the bit. Now the data centers themselves have the scale to drive >> Absolutely. >> the scale innovation that before was before was really limited to either a PC or a phone or some more consumer device. >> Absolutely the case. When you take that cross-section of hard drive applications, that's a hundred percent the case, and in fact, we look at the utilization as a vertically integrated company we look at our media facilities for the disks, we look at our wafer facilities for heads, and those facilities as we look forward are going to be as busy as busier than they've ever been. I mean the amount of data is relative to the density as well as disks and heads and how many you can employ. So we see this in terms of fundamental technology and component construction, manufacturing, busier than it's ever been. We'll make fewer units. I mean there will be fewer units as they become bigger and denser for this application space, but the fundamental consumption of magnetic recording technology and components is all-time records. >> Right. And you haven't even talked about the software-defined piece that's dragging the utilization of that data across multiple applications. >> And it's just one of these that come in to help everybody there too, yeah. >> Jeff: You got another 35 years more years in you? (both laugh) >> I hope so. >> All right. >> But that would be the edge of it, I think. >> All right, we're going to take Mark Grace here, only 35 more years, Lord knows what he'll be working on. Well Mark, thanks for taking a few minutes and answering your prospective >> No that's fine, thanks a lot. >> Absolutely, Mark Grace, I'm Jeff Frick, you're watching theCUBE from Western Digital headquarters in San Jose, California. Thanks for watching. >> Mark: All right.

(upbeat techno music) >> Announcer: Live from San Jose, California it's theCUBE, covering Innovating to Fuel the Next Decade of Big Data, brought to you by Western Digital. >> Hey, welcome back everybody. Jeff Frick here at theCUBE. We're at the Western Digital Headquarters off Almaden down in San Jose, a really important place. Western Digital's been here for a while, their headquarters. A lot of innovation's been going on here forever. So we're excited to be here really for the next generation. The event's called Innovating to Fuel the Next Generation of big data, and we're joined by many time Cuber, Dave Tang. He is the SVP in corporate marketing from Western Digital. Dave, always great to see you. >> Yeah. Always great to be here, Jeff. Thanks. >> Absolutely. So you got to MC the announcement today. >> Yes. >> So for the people that weren't there, let's give them a quick overview on what the announcement was and then we can dive in a little deeper. >> Great, so what we were announcing was a major breakthrough in technology that's going to allow us to drive the increase in capacity in density to support big data for the next decade and beyond, right? So capacities and densities had been starting to level off in terms of hard drive technology capability. So what we announced was microwave-assisted magnetic recording technology that will allow us to keep growing that areal density up and reducing the cost per terabyte. >> You know, it's fascinating cause everyone loves to talk about Moore's Law and have these silly architectural debates, whether Moore's Law is alive or dead, but, as anyone who's lived here knows, Moore's Law is really an attitude much more it is than the specific physics of microprocessor density growth. And it's interesting to see. As we know the growth of data is growing in giant and the types of data, and not only regular big data, but now streaming data are bigger and bigger and bigger. I think you talked about stuff coming off of people and machines compared to business data is way bigger. >> Right. >> But you guys continue to push limits and break through, and even though we expect everything to be cheaper, faster, and better, you guys actually have to execute it-- >> Dave: Right. >> Back at the factory. >> Right, well it's interesting. There's this healthy tension, right, a push and pull in the environment. So you're right, it's not just Moore's Law that's enabling a technology push, but we have this virtuous cycle, right? We've realized what the value is of data and how to extract the possibilities and value of data, so that means that we want to store more of that data and access more of that data, which drives the need for innovation to be able to support all of that in a cost effective way. But then that triggers another wave of new applications, new ways to tap into the possibilities of data. So it just feeds on itself, and fortunately we have great technologists, great means of innovation, and a great attitude and spirit of innovation to help drive that. >> Yeah, so for people that want more, they can go to the press releases and get the data. We won't dive deep into the weeds here on the technology, but I thought you had Janet George speak, and she's chief data scientist. Phenomenal, phenomenal big brain. >> Dave: Yes. >> A smart lady. But she talked about, from her perspective we're still just barely even getting onto this data opportunity in terms of automation, and we see over and over at theCUBE events, innovation's really not that complicated. Give more people access to the data, give them more access to the tools, and let them try things easier and faster and feel quick, there's actually a ton of innovation that companies can unlock within their own four walls. But the data is such an important piece of it, and there's more and more and more of this. >> Dave: Right, right. >> What used to be digital exhaust now is, I think maybe you said, or maybe it was Dave, that there's a whole economy now built on data like we used to do with petroleum. I thought that was really insightful. >> Yeah, right. It's like a gold mine. So not only are the sources of data increasing, which is driving increased volume, but, as Janet was alluding to, we're starting to come up with the tools and the sophistication with machine learning and artificial intelligence to be able to put that data to new use as well as to find the pieces of data to interconnect, to drive these new capabilities and new insights. >> Yeah, but unlike petroleum it doesn't get used up. I mean that's the beauty of data. (laughing) >> Yeah, that's right. >> It's a digital process that can be used over and over and over again. >> And a self-renewing, renewing resource. And you're right, in that sense that it's being used over and over again that the longevity of that data, the use for life is growing exponentially along with the volume. >> Right, and Western Digital's in a unique position cause you have systems and you have big systems that could be used in data centers, but you also have the media that powers a whole bunch of other people's systems. So I thought one of the real important announcements today was, yes it's an interesting new breakthrough technology that uses energy assist to get more density on the drives, but it's done in such a way that the stuff's all backward compatible. It's plug and play. You've got production scheduled in a couple years I think with test out the customers-- >> Dave: That's right. >> Next year. So, you know, that is such an important piece beyond the technology. What's the commercial acceptance? What are the commercial barriers? And this sounds like a pretty interesting way to skin that cow. >> Right, often times the best answers aren't the most complex answers. They're the more elegant and simplistic answers. So it goes from the standpoint of a user being able to plug and play with older systems, older technologies. That's beautiful, and for us, to be able to, the ability to manufacture it in high volume reliably and cost effectively is equally as important. >> And you also talked, which I think was interesting, is kind of the relationship between hard drives and flash, because, obviously, flash is a, I want to say the sexy new toy, but it's not a sexy new toy anymore. >> Right. >> It's been around for a while, but, with that pressure on flash performance, you're still seeing the massive amounts of big data, which is growing faster than that. And there is a rule for the high density hard drives in that environment, and, based on the forecast you shared, which I'm presuming came from IDC or people that do numbers for a living, still a significant portion of a whole lot of data is not going to be on flash. >> Yeah, that's right. I think we have a tendency, especially in technology, to think either or, right? Something is going to take over from something else, but in this case it's definitely an and, right. And a lot of that is driven by this notion that there's fast data and big data, and, while our attention seems to shift over to maybe some fast data applications like autonomous vehicles and realtime applications, surveillance applications, there's still a need for big data because the algorithms that drive those realtime applications have to come from analysis of vast amounts of data. So big data is here to stay. It's not going away or shifting over. >> I think it's a really interesting kind of cross over, which Janet talked about too, where you need the algorithms to continue sharing the system that are feeding, continuing, and reacting to the real data, but then that just adds more vocabulary to their learning set so they can continue to evolve overtime. >> Yeah, what really helps us out in the market place is that because we have technologies and products across that full spectrum of flash and rotating magnetic recording, and we sell to customers who buy devices as well as platforms and systems, we see a lot of applications, a lot of uses of data, and we're able to then anticipate what those needs are going to be in the near future and in the distant future. >> Right, so we're getting towards the end of 2017, which I find hard to say, but as you look forward kind of to 2018 and this insatiable desire for more storage, cause this insatiable creation of more data, what are some of your priorities for 2018 and what are you kind of looking at as, like I said, I can't believe we're going to actually flip the calendar here-- >> Dave: Right, right. >> In just a few short months. (laughing) >> Well, I think for us, it's the realization that all these applications that are coming at us are more and more diverse, and their needs are very specialized. So it's not just the storage, although we're thought of as a storage company, it's not just about the storage of that data, but you have contrive complete environments to capture and preserve and access and transform that data, which means we have to go well beyond storage and think about how that data is accessed, technical interfaces to our memory products as well as storage products, and then where compute sits. Does it still sit in a centralized place or do you move compute to out closer to where the data sits. So, all this innovation and changing the way that we think about how we can mine that data is top of the mind for us for the next year and beyond. >> It's only job security for you, Dave. (laughing) >> Dave: Funny to think about. >> Alright. He's Dave Tang. Thanks for inviting us and again congratulations on the presentation. >> Always a pleasure. >> Alright, Dave Tang, I'm Jeff Frick. You're watching theCUBE from Western Digital headquarters in San Jose, California. Thanks for watching. (upbeat techno music)

>> Announcer: Live from San Jose, California, it's The Cube. Covering Innovating to Fuel the Next Decade of Big Data. Brought to you by Western Digital. >> Hey, welcome back everybody. Jeff Frick here with The Cube. We're at the Western Digital headquarters in San Jose, the Great Oaks Campus, a really historic place in the history of Silicon Valley and computing. It's The Innovating to Fuel the Next Generation of Big Data event with Western Digital. We're really excited to be joined by our next guest, Mike Cordano. He's the president and chief operating officer of Western Digital. Mike, great to see you. >> Great to see you as well. Happy you guys could be here. It's an exciting day. >> Absolutely. First off, I think the whole merger thing is about done, right? That's got to feel good. >> Yeah, it's done, but there's legs to it, right? So we've combined these companies now, three of them, three large ones, so obviously Western Digital and Hitachi Global Storage, now we've added SanDisk into one Western Digital, so we're all together. Obviously more to do, as you expect in a large scale integration. There will be a year or two of bringing all those business processes and systems together, but I got to say, the teams are coming together great, showing up in our financial performance and our product execution, so things are really coming together. >> Yeah, not an easy task by any stretch of the imagination. >> No, not easy, but certainly a compliment to our team. I mean, we've got great people. You know, like anything, if you can harness the capabilities of your team, there's a lot you can accomplish, and it really is a compliment to the team. >> Excellent. Well, congratulations on that, and talking a bit about this event here today, you've even used "Big Data" in the title of the event, so you guys are obviously in a really unique place, Western Digital. You make systems, big systems. You also make the media that feeds a lot of other people's systems, but as the big data grows, the demand for data grows, it's got to live somewhere, so you're sitting right at the edge where this stuff's got to sit. >> Yeah, that's right, and it's central to our strategy, right? So if you think about it, there's three fundamental technologies that we think are just inherent in all of the evolution of compute and IT architecture. Obviously, there is compute, there is storage or memory, and then there's sort of movement, or interconnect. We obviously live in the storage or memory node, and we have a very broad set of capabilities, all the way from rotating magnetic media, which was our heritage, now including non-volatile memory and flash, and that's just foundational to everything that is going to come, and as you said, we're not going to stop there. It's not just a devices or component company, we're going to continue to innovate above that into platforms and systems, and why that becomes important to us, is there's a lot of technology innovation we can do that enhances the offering that we can bring to market when we control the entire technology stat. >> Right. Now, we've had some other guests on and people can get more information on the nitty-gritty details of the announcement today, the main announcement. Basically, in a nutshell, enabling you to get a lot more capacity in hard drives. But, I thought in your opening remarks this morning, there were some more high-level things I wanted to dig into with you, and specifically, you made an analogy of the data economy, and compared it to the petroleum economy. I've never... A lot of times, they talk about big data, but no one really talks about it, that I've heard, in those terms, because when you think about the petroleum economy, it's so much more than fuel and cars, and the second-order impacts, and the third-order impacts on society are tremendous, and you're basically saying, "We're going to "do this all over again, but now it's based on data." >> Yeah, that's right, and I think it puts it into a form that people can understand, right? I think it's well-proven what happened around petroleum, so the discovery of petroleum, and then the derivative industries, whether it be automobiles, whether it be plastics, you pick it, the entire economy revolved around, and, to some degree, still revolves around petroleum. The same thing will occur around data. You're seeing it with investments, you hear now things like machine learning, or artificial intelligence, that is all ways to transform and mine data to create value. >> Right. >> And we're going to see industries change rapidly. Autonomous cars, that's going to be enabled by data, and capabilities here, so pick your domain. There's going to be innovation across a lot of fronts, across a lot of traditional vertical industries, that is all going to be about data and driven by data. >> It's interesting what Janet, Doctor Janet George talked about too a little bit is the types of data, and the nozzles of the data is also evolving very quickly from data at rest to data in motion, to real-time analytics, to, like you said, the machine learning and the AI, which is based on modeling prior data, but then ingesting new data, and adjusting those models so even the types and the rate and the speed of the data is under dramatic change right now. >> Yeah, that's right, and I think one of the things that we're helping enable is you kind of get to this concept of what do you need to do to do what you describe? There has to be an infrastructure there that actually enables it. So, when you think about the scale of data we're dealing with, that's one thing that we're innovating around, then the issue is, how do you allow multiple applications to simultaneously access and update and transform that? Those are all problems that need to be solved in the infrastructure to enable things like AI, right? And so, where we come into play, is creating that infrastructure layer that actually makes that possible. The other thing I talked about briefly in the Q and A was, think about the problem of a future where the data set is just too large to actually move it in a substantive way to the compute. We actually have to invert that model over time architecturally, and bring the compute to the data, right? Because it becomes too complicated and too expensive to move from the storage layer up to compute and back, right? That is a complex operation. That's why those three pillars of technology are so important. >> And you've talked, and we're seeing in the Cloud right, because this continuing kind of atomization, atomic, not automatic, but making these more atomic. A smaller unit that the Cloud has really popularized, so you need a lot, you need a little, really, by having smaller bits and bytes, it makes that that much more easy. But another concept that you delved into a little was fast data versus big data, and clearly flash has been the bright, shiny object for the last couple years, and you guys play in that market as well, but it is two very different ways to think of the data, and I thought the other statistic that was shared is you know, the amount of data coming off of the machines and people dwarfs the business data, which has been the driver of IT spend for the last several decades. >> Yeah, no, that's right, and sort of that... You think about that, and the best analogy is a broader definition of IOT, right? Where you've got all of these censors, whether it be camera censors, because that's just a censor, creating an image or a video, or if it's more industrialized too, you've got all these sources of data, and they're going to proliferate at an exponential rate, and our ability to aggregate that in some sort of an organized way, and then act upon it, again, let's use the autonomous car as the example. You've got all these censors that are in constant motion. You've got to be able to aggregate the data, and make decisions on it at the edge, so that's not something... You can't deal with latency up to the Cloud and back, if it's an automobile, and it needs to make an instantaneous decision, so you've got to create that capability locally, and so when you think about the evolution of all this, it's really the integration of the Cloud, which, as Janet talked about, is the ability to tap into this historical or legacy data to help inform a decision, but then there's things happening out at the edge that are real time, and you have to have the capability to ingest the content, make a decision on it very quickly, and then act on it. >> Right. There's a great example. We went to the autonomous... Just navigation for the autonomous vehicles. It's own subset that I think Goldman-Sachs said it a seven billion dollar industry in the not-too-distant future, and the great example is this combination of the big data and the live data is, when they actually are working on the road. So you've got maps that tell you, and are updated, kind of what the road looks like, but on Tuesday, they were shifting the lane, and that particular lane now has cones in it, so the combination of the two is such a powerful thing. >> That's right. >> I want to dive into another topic we talked about, which is really architecting for the future. Unlike oil, data doesn't get consumed and is no longer available, right? It's a reusable asset, and you talked about classic stove-topping of data within an application center world where now you want that data available for multiple applications, so very different architecture to be able to use it across many fronts, some of which you don't even know yet. >> That's right. I think that's a key point. One of the things, when we talk to CEOs, or CIOs I should say, what they're realizing, to the extent you can enable a cost-effective mechanism for me to store and keep everything, I don't know how I'll derive value from it some time in the future, because as applications evolve, we're finding new insights into what can help drive decisions or innovation, or, to take it to health care, some sort of innovation that cures disease. That's one of the things that everybody wants to do. I want to build aggregate everything. If I could do that cost effectively enough, I'll find a way to get value out of it over time, and that's something where, when we're thinking about big data and what we talked about today, that's central to that idea, and enabling it. >> Right, and digital transformation, right, the hot buzz word, but we hear, time and time again, such a big piece of that is giving the democratization. Democratization of the data, so more people have access to it, democratization of the tools to manipulate that data, not just Mahogany Row super smart people, and then to have a culture that lets people actually try, experiment, fail fast, and there's a lot of innovation that would be unlocked right within your four walls, that probably are not being tapped into. >> Well, that's right, and that's something that innovation, and an innovation culture is something that we're working hard at, right? So if you think about Western Digital, you might think of us as, you know, legacy Western Digital as sort of a fast following, very operational-centric company. We're still good at those things, but over the last five years, we've really pushed this notion of innovation, and really sort of pressing in to becoming more influential in those feature architectures. That drives a culture that, if we think about the technical community, if we create the right sort of mix of opportunity, appetite for some risk, that allows the best creativity to come out of our technical... Innovating along these lines. >> Right, I'll give you the last word. I can't believe we're going to turn the calendar here on 2017, which is a little scary. As you look forward to 2018, what are some of your top priorities? What are you going to be working on as we come into the new calendar year? >> Yeah, so as we look into 2018 and beyond, we really want to drive this continued architectural shift. You'll see us be very active, and I think you talked about it, you'll see us getting increasingly active in this democratization. So we're going to have to figure out how we engage the broader open-source development world, whether it be hardware or software. We agree with that mantra, we will support that. Obviously we can do unique development, but with some hooks and keys that we can drive a broader ecosystem movement, so that's something that's central to us, and one last word would be, one of the things that Martin Fink has talked about which is really part of our plans as we go onto the new year, is really this inverting the model, where we want to continue to drive an architecture that brings compute to the storage and enables some things that just can't be done today. >> All right, well Mike Cordano, thanks for taking a few minutes, and congratulations on the terrific event. >> Thank you. Appreciate it. >> He's Mike Cordano, I'm Jeff Frick, you're watching The Cube, we're at Western Digital headquarters in San Jose, Great Oaks Campus, it's historic. Check it out. Thanks for watching.

>> Male voiceover: Live from San Jose California, it's the Cube, covering Innovating to Fuel the Next Decade of Big Data. Brought to you by Western Digital. >> Hey, welcome back everybody, Jeff Frick here with the Cube. We're at the Western Digital World Headquarters It's the Almaden Campus in San Jose. If you know anything about the tech world, you know there's a lot of innovation that's been happening on this campus for years and years and years. Big announcement today called Innovating to Fuel the Next Generation of Big Data. Lot of exciting announcements and here to join us to tell us all about it is Brendan Collins. He's the Vice President of Product Marketing Devices for Western Digital. Brendan, great to see you. >> Thank you, glad to be here. >> Absolutely so, really exciting announcement. You know, I've talked to Kim Stevenson at Intel, we had an interview talking about Moore's law. And one thing she really reinforced is that Moore's law is really more of an attitude than it is specifically physics, and whether you want to argue the physics is one thing, but the attitude for innovation, to continue to deliver a lot more for less, just continues, continues, and continues, and you guys announced a huge step in that direction today. >> Yeah, we have a challenge that storage is growing at a rate of about 40 percent per year. And budgets from the data centers are not growing, right? So the challenge is for us to develop new technologies that allow us to stay on the technology curve, and cut costs and do that efficiently. >> Then this is a big one, so let's jump in. So actually it was years ago I was actually at the event when you guys introduced the Helium drives, and that was a big deal there, and you've continued to kind of move that innovation but then you can see a plateau. And the density of this data, so you guys had to come up with something new. >> Yeah, what we've seen is that our PMR technology that we use currently is slowly running out of steam, right? So in order to come down the cost curve, we needed to boost areal density. And luckily we were able to come up with a new breakthrough in MAMR technology that will allow us to do that for the next decade. >> It's interesting in the talk, you talked about you guys could see this kind of coming and you actually put a lot of bets on the table, you didn't just bet on MAMR, you bet on HAMR, and you continued along a number of multiple tracks, and you've been at this for a while. What was kind of the innovation that finally gave you a breakthrough moment that got us to where we are today? >> Well, there were multiple technologies that we could have invested in, and we decided to continue on the two major ones which were HAMR and MAMR but we made a decision to invest in a process called, a head fabrication process called damascene that allowed us to extend the life of PMR for the last five to six years, and it's been in all the products we've been shipping since 2013. >> And you talked the areal density, so that's basically the amount of information we can put on the square inch of surface area And you've really, you attacked it on two vectors. One is how many tracks, just think of a record, how many tracks can you get on an album, in terms of the number of lines, and then how much density then you can have on each of those tracks. >> That's right, that's right. And you're now seeing major improvements on both of those factors. >> Well if you look at, we've had three enabling technologies in our products for the past three to four years, right. One is helium, one is micro actuation, and the other is the damascene process. Damascene and micro actuation actually push track density which enables higher capacity. But the newer technology that we're talking about, MAMR, addresses both factors. So we push the track density even tighter together, But we also boost the linear density at the same time, and we do that without adding cost. >> Right. The other thing you talked about, and I think it's a really important piece, right it's not only the technology breakthrough, but it's also how does that fit within the existing ecosystem of your customers, and obviously big giant data centers and big giant cloud providers, we actually have a show going on at a big cloud show right now, and this technology was innovative in that you've got a breakthrough on density, but not so crazy that you introduced a whole bunch of new factors into the ecosystem that would then have to be incorporated into all these systems, because you guys not only make your own systems, but you make the media that feeds a whole host of ecosystems, and that was a pretty important piece. >> If you look at some previous technologies we've introduced whether it be even 4K sectors in the industry, or shingled magnetic reporting, both of those require whole side modifications. Any time you have whole side modifications, it generally slows down the adoption, right? With HAMR, one of the challenges that we had was because of the concerns with thermals on the media, we needed a process called wear leveling, and that required whole software changes. In contrast, when we go to MAMR, everything is seamless, everything is transparent, and it's great. >> Right. I thought it was much simpler than that. I thought just heat is bad, HAMR is heat, and MAMR is microwave, and you know, heat and efficiency and data centers and all those, kind of again, system-level concerns; heat's never a good thing in electronics. >> Well, and in the case of MAMR versus HAMR, there's like an order of magnitude difference in the temperature on the disk, which is the key concern. >> And then of course as you mentioned in the key note, this is real, you've got sample units going on, correct me if I'm wrong, as early as next year >> That's right. >> you're hoping you'd be in scale production in 2020. Where some of these other competing technologies, there's really still no forecasted ship date on the horizon. >> Yeah, you can generate samples, you can build lower quantities of these HAMR drives, but you still have that big concern out there in front of you, how do I address the reliability, how do I address the complexity of all these new materials, and then if I got all of that to work, how do I do it commercially because of the cost additives. >> Right; so I just want to get your perspective before we let you go, you're busy, there's a high demand for your time, as you kind of think back and look at these increasing demands for storage, this increasing demand for computers, and I think one of the data points given is, you know, the data required for humans and machines and IOT is growing way way way way faster than business focused data which has been the driver of a lot of this stuff, if you just kind of sit back and take a look, you know, what are some of your thoughts because I'm sure not that long ago you could have never imagined that there would be the demand for the types of capacities that we're talking about now and we both know that when we sit down five years from now, ten years ago, you know, ten years from now, we're going to look back at today and think, you know, that was zero. >> Yeah, way back in the day there were just PCs and servers and there was traditional IT with rate, today with autonomous cars and IOT and AI and machine learning, it's just going to continue, so that exponential growth that you saw, there's no sign of that slowing down, which is good news for us. >> Yeah, good job security for you for sure. >> You bet! >> Alright Brendan, well, again, thanks for taking a few minutes to sit down and congratulations on the great event and the launch of these new products. >> Thank you, thank you. >> He's Brendan Collins, I'm Jeff Frick, you're watching the Cube from the Western Digital Headquarters in San Jose California. Thanks for watching.

>> Presenter: From the Fairmont Hotel, in the heart of Silicon Valley, it's theCUBE. Covering When IoT Met AI The Intelligence of Things. Brought to you by Western Digital. >> Hey welcome back everybody, Jeff Frick here with theCUBE. We're in downtown San Jose at the Fairmont Hotel, at an event called When IoT Met AI The Intelligence of Things. You've heard about the internet of things, and on the intelligence of things, it's IoT, it's AI, it's AR, all this stuff is really coming to play, it's very interesting space, still a lot of start-up activity, still a lot of big companies making plays in this space. So we're excited to be here, and really joined by our host, big thanks to Western Digital for hosting this event with WDLabs' Dave Tang. Got newly promoted since last we spoke. The SVP of corporate marketing and communications, for Western Digital, Dave great to see you as usual. >> Well, great to be here, thanks. >> So I don't think the need for more storage is going down anytime soon, that's kind of my takeaway. >> No, no, yeah. If this wall of data just keeps growing. >> Yeah, I think the term we had yesterday at the Ag event that we were at, also sponsored by you, is really the flood of data using an agricultural term. But it's pretty fascinating, as more, and more, and more data is not only coming off the sensors, but coming off the people, and used in so many more ways. >> That's right, yeah we see it as a virtual cycle, you create more data, you find more uses for that data to harness the power and unleash the promise of that data, and then you create even more data. So, when that virtual cycle of creating more, and finding more uses of it, and yeah one of the things that we find interesting, that's related to this event with IoT and AI, is this notion that data is falling into two general categories. There's big data, and there's fast data. So, big data I think everyone is quite familiar with by this time, these large aggregated likes of data that you can extract information out of. Look for insights and connections between data, predict the future, and create more prescriptive recommendations, right? >> Right. >> And through all of that you can gain algorithms that help to make predictions, or can help machines run based on that data. So we've gone through this phase where we focused a lot on how we harness big data, but now we're taking these algorithms that we've gleaned from that, and we're able to put them in real time applications, and that's sort of been the birth of fast data, it's been really-- >> Right, the streaming data. We cover Spark Summit, we cover Flink, and New, a new kind of open source project that came out of Berlin. That some people would say the next generation of Spark, and the other thing, you know, good for you guys, is that it used to be, not only was it old data, but it was a sampling of old data. Now on this new data, and the data stream that's all of the data. And I would actually challenge, I wonder if that separation as you describe, will stay, because I got to tell you, the last little drive I bought, just last week, was an SSD drive, you know, one terabyte. I needed some storage, and I had a choice between spinning disc and not, and I went with the flat. I mean, 'cause what's fascinating to me, is the second order benefits that we keep hearing time, and time, and time again, once people become a data-driven enterprise, are way more than just that kind of top-level thing that they thought. >> Exactly, and that's sort of that virtual cycle, you got to taste, and you learn how to use it, and then you want more. >> Jeff: Right, right. >> And that's the great thing about the breadth of technologies and products that Western Digital has, is from the solid state products, the higher performance flash products that we have, to the higher capacity helium-filled drive technologies, as well as devices going on up into systems, we cover this whole spectrum of fast data and big data. >> Right, right. >> I'll give an example. So credit card fraud detection is an interesting area. Billions of dollars potentially being lost there. Well to learn how to predict when transactions are fraudulent, you have to study massive amounts of data. Billions of transactions, so that's the big data side of it, and then as soon as you do that, you can take those algorithms and run them in real time. So as transactions come in for authorization, those algorithms can determine, before they're approved, that one's fraudulent, and that one's not. Save a lot of time and processing for fraud claims. So that's a great example of once you learn something from big data, you apply it to the real-time realm, and it's quite dire right? And then that spawned you to collect even more data, because you want to find new applications and new uses. >> Right, and too kind of this wave of computing back and forth from the shared services computer, then the desktop computer, now it's back to the cloud, and then now it's-- >> Dave: Out with the edge. >> IoT, it's all about the edge. >> Yeah, right. >> And at the end of the day, it's going to be application-specific. What needs to be processed locally, what needs to be processed back at the computer, and then all the different platforms. We were again at a navigation for autonomous vehicles show, who knew there was such a thing that small? And even the attributes of the storage required in the ecosystem of a car, right? And the environmental conditions-- >> That's right. >> Is the word I'm looking for. Completely different, new opportunity, kind of new class of hardware required to operate in that environment, and again that still combines cloud and Edge, sensors and maps. So just the, I don't think that the man's going down David. >> Yeah, absolutely >> I think you're in a good spot. (Jeff laughing) >> You're absolutely right, and even though we try to simplify into fast data, and big data, and Core and Edge, what we're finding is that applications are increasingly specialized, and have specialized needs in terms of the type of data. Is it large amounts of data, is it streaming? You know, what are the performance characteristics, and how is it being transformed, what's the compute aspect of it? And what we're finding, is that the days of general-purpose compute and storage, and memory platforms, are fading, and we're getting into environments with increasingly specialized architectures, across all those elements. Compute, memory and storage. So that's what's really exciting to be in our spot in the industry, is that we're looking at creating the future by developing new technologies that continue to fuel that growth even further, and fuel the uses of data even further. >> And fascinating just the ongoing case of Moore's law, which I know is not, you know you're not making microprocessors, but I think it's so powerful. Moore's law really is a philosophy, as opposed to an architectural spec. Just this relentless pace of innovation, and you guys just continue to push the envelope. So what are your kind of priorities? I can't believe we're halfway through 2017 already, but for kind of the balance of the year kind of, what are some of your top-of-mind things? I know it's exciting times, you're going through the merger, you know, the company is in a great space. What are your kind of top priorities for the next several months? >> Well, so, I think as a company that has gone through serial acquisitions and integrations, of course we're continuing to drive the transformation of the overall business. >> But the fun stuff right? It's not to increase your staff (Jeff laughing). >> Right, yeah, that is the hardware. >> Stitching together the European systems. >> But yeah, the fun stuff includes pushing the limits even further with solid state technologies, with our 3D NAND technologies. You know, we're leading the industry in 64 layer 3D NAND, and just yesterday we announced a 96 layer 3D NAND. So pushing those limits even further, so that we can provide higher capacities in smaller footprints, lower power, in mobile devices and out on the Edge, to drive all these exciting opportunities in IoT an AI. >> It's crazy, it's crazy. >> Yeah it is, yeah. >> You know, terabyte SD cards, terabyte Micro SD cards, I mean the amount of power that you guys pack into these smaller and smaller packages, it's magical. I mean it's absolutely magic. >> Yeah, and the same goes on the other end of the spectrum, with high-capacity devices. Our helium-filled drives are getting higher and higher capacity, 10, 12, 14 terabyte high-capacity devices for that big data core, that all the data has to end up with at some point. So we're trying to keep a balance of pushing the limits on both ends. >> Alright, well Dave, thanks for taking a few minutes out of your busy day, and congratulations on all your success. >> Great, good to be here. >> Alright, he's Dave Tang from Western Digital, he's changing your world, my world, and everyone else's. We're here in San Jose, you're watching theCUBE, thanks for watching.

(upbeat electronic music) >> Narrator: From the Fairmont Hotel in the heart of Silicon Valley, it's theCUBE. Covering when IoT met AI, The Intelligence of Things. Brought to you by Western Digital. >> Welcome back here everybody, Jeff Frick here with theCUBE. We are at downtown San Jose at the Fairmont Hotel. When IoT met AI it happened right here, you saw it first. The Intelligence of Things, a really interesting event put on by readwrite and Western Digital and we are really excited to welcome back a many time CUBE alumni and always a fan favorite, she's Janet George. She's Fellow & Chief Data Officer of Western Digital. Janet, great to see you. >> Thank you, thank you. >> So, as I asked you when you sat down, you're always working on cool things. You're always kind of at the cutting edge. So, what have you been playing with lately? >> Lately I have been working on neural networks and TensorFlow. So really trying to study and understand the behaviors and patterns of neural networks, how they work and then unleashing our data at it. So trying to figure out how it's training through our data, how many nets there are, and then trying to figure out what results it's coming with. What are the predictions? Looking at how the predictions are, whether the predictions are accurate or less accurate and then validating the predictions to make it more accurate, and so on and so forth. >> So it's interesting. It's a different tool, so you're learning the tool itself. >> Yes. >> And you're learning the underlying technology behind the tool. >> Yes. >> And then testing it actually against some of the other tools that you guys have, I mean obviously you guys have been doing- >> That's right. >> Mean time between failure analysis for a long long time. >> That's right, that's right. >> So, first off, kind of experience with the tool, how is it different? >> So with machine learning, fundamentally we have to go into feature extraction. So you have to figure out all the features and then you use the features for predictions. With neural networks you can throw all the raw data at it. It's in fact data-agnostic. So you don't have to spend enormous amounts of time trying to detect the features. Like for example, If you throw hundreds of cat images at the neural network, the neural network will figure out image features of the cat; the nose, the eyes, the ears and so on and so forth. And once it trains itself through a series of iterations, you can throw a lot of deranged cats at the neural network and it's still going to figure out what the features of a real cat is. >> Right. >> And it will predict the cat correctly. >> Right. So then, how does that apply to, you know, the more specific use case in terms of your failure analysis? >> Yeah. So we have failures and we have multiple failures. Some failures through through the human eye, it's very obvious, right? But humans get tired, and over a period of time we can't endure looking at hundreds and millions of failures, right? And some failures are interconnected. So there is a relationship between these failure patterns or there is a correlation between two failures, right? It could be an edge failure. It could a radial failure, eye pattern type failure. It could be a radial failure. So these failures, for us as humans, we can't escape. >> Right. >> And we used to be able to take these failures and train them at scale and then predict. Now with neural networks, we don't have to take and do all that. We don't have to extract these labels and try to show them what these failures look like. Training is almost like throwing a lot of data at the neural networks. >> So it almost sounds like kind of the promise of the data lake if you will. >> Yes. >> If you have heard about, from the Hadoop Summit- >> Yes, yes, yes. >> For ever and ever and ever. Right? You dump it all in and insights will flow. But we found, often, that that's not true. You need hypothesis. >> Yes, yes. >> You need to structure and get it going. But what you're describing though, sounds much more along kind of that vision. >> Yes, very much so. Now, the only caveat is you need some labels, right? If there is no label on the failure data, it's very difficult for the neural networks to figure out what the failure is. >> Jeff: Right. >> So you have to give it some labels to understand what patterns it should learn. >> Right. >> Right, and that is where the domain experts come in. So we train it with labeled data. So if you are training with a cat, you know the features of a cat, right? In the industrial world, cat is really what's in the heads of people. The domain knowledge is not so authoritative. Like the sky or the animals or the cat. >> Jeff: Right. >> The domain knowledge is much more embedded in the brains of the people who are working. And so we have to extract that domain knowledge into labels. And then you're able to scale the domain. >> Jeff: Right. >> Through the neural network. >> So okay so then how does it then compare with the other tools that you've used in the past? In terms of, obviously the process is very different, but in terms of just pure performance? What are you finding? >> So we are finding very good performance and actually we are finding very good accuracy. Right? So once it's trained, and it's doing very well on the failure patterns, it's getting it right 90% of the time, right? >> Really? >> Yes, but in a machine learning program, what happens is sometimes the model is over-fitted or it's under-fitted or there is bias in the model and you got to remove the bias in the model or you got to figure out, well, is the model false-positive or false-negative? You got to optimize for something, right? >> Right, right. >> Because we are really dealing with mathematical approximation, we are not dealing with preciseness, we are not dealing with exactness. >> Right, right. >> In neural networks, actually, it's pretty good, because it's actually always dealing with accuracy. It's not dealing with precision, right? So it's accurate most of the time. >> Interesting, because that's often what's common about the kind of difference between computer science and statistics, right? >> Yes. >> Computers is binary. Statistics always has a kind of a confidence interval. But what you're describing, it sounds like the confidence is tightening up to such a degree that it's almost reaching binary. >> Yeah, yeah, exactly. And see, brute force is good when your traditional computing programing paradigm is very brute force type paradigm, right? The traditional paradigm is very good when the problems are simpler. But when the problems are of scale, like you're talking 70 petabytes of data or you're talking 70 billion roles, right? Find all these patterns in that, right? >> Jeff: Right. >> I mean you just, the scale at which that operates and at the scale at which traditional machine learning even works is quite different from how neural networks work. >> Jeff: Okay. >> Right? Traditional machine learning you still have to do some feature extraction. You still have to say "Oh I can't." Otherwise you are going to have dimensionality issues, right? It's too broad to get the prediction anywhere close. >> Right. >> Right? And so you want to reduce the dimensionality to get a better prediction. But here you don't have to worry about dimensionality. You just have to make sure the labels are right. >> Right, right. So as you dig deeper into this tool and expose all these new capabilities, what do you look forward to? What can you do that you couldn't do before? >> It's interesting because it's grossly underestimating the human brain, right? The human brain is supremely powerful in all aspects, right? And there is a great deal of difficulty in trying to code the human brain, right? But with neural networks and because of the various propagation layers and the ability to move through these networks we are coming closer and closer, right? So one example: When you think about driving, recently, Google driverless car got into an accident, right? And where it got into an accident was the driverless car was merging into a lane and there was a bus and it collided with the bus. So where did A.I. go wrong? Now if you train an A.I., birds can fly, and then you say penguin is a bird, it is going to assume penguin can fly. >> Jeff: Right, right. >> We as humans know penguin is a bird but it can't fly like other birds, right? >> Jeff: Right. >> It's that anomaly thing, right? Naturally when are driving and a bus shows up, even if it's yield, the bus goes. >> Jeff: Right, right. >> We yield to the bus because it's bigger and we know that. >> A.I. doesn't know that. It was taught that yield is yield. >> Right, right. >> So it collided with the bus. But the beauty is now large fleets of cars can learn very quickly based on what it just got from that one car. >> Right, right. >> So now there are pros and cons. So think about you driving down Highway 85 and there is a collision, it's Sunday morning, you don't know about the collision. You're coming down on the hill, right? Blind corner and boom that's how these crashes happen and so many people died, right? If you were driving a driverless car, you would have knowledge from the fleet and from everywhere else. >> Right. >> So you know ahead of time. We don't talk to each other when we are in cars. We don't have universal knowledge, right? >> Car-to-car communication. >> Car-to-car communications and A.I. has that so directly it can save accidents. It can save people from dying, right? But people still feel, it's a psychology thing, people still feel very unsafe in a driverless car, right? So we have to get over- >> Well they will get over that. They feel plenty safe in a driverless airplane, right? >> That's right. Or in a driveless light rail. >> Jeff: Right. >> Or, you know, when somebody else is driving they're fine with the driver who's driving. You just sit in the driver's car. >> But there's that one pesky autonomous car problem, when the pedestrian won't go. >> Yeah. >> And the car is stopped it's like a friendly battle-lock. >> That's right, that's right. >> Well good stuff Janet and always great to see you. I'm sure we will see you very shortly 'cause you are at all the great big data conferences. >> Thank you. >> Thanks for taking a few minutes out of your day. >> Thank you. >> Alright she is Janet George, she is the smartest lady at Western Digital, perhaps in Silicon Valley. We're not sure but we feel pretty confident. I am Jeff Frick and you're watching theCUBE from When IoT meets AI: The Intelligence of Things. We will be right back after this short break. Thanks for watching. (upbeat electronic music)

>> They are disruptive technologies. And if you think about the disruption that's happening in business, with IoT, with OT, and with big data, you can't get anything more disruptive to the whole of the business chain as this particular area. It's an area that I focused on myself, asking the question, should everything go to the cloud? Is that the new future? Is 90% of the computing going to go to the cloud with just little mobile devices right on the edge? Felt wrong when I did the math on it, I did some examples of real-world environments, wind farms, et cetera, it clearly was not the right answer, things need to be near the edge. And I think one of the areas to me that solidified it was when you looked at an area like video. Huge amounts of data, real important decisions being made on the content of that video, for example, recognizing a face, a white hat or a black hat. If you look at the technology, sending that data somewhere to do that recognition just does not make sense. Where is it going? It's going actually into the camera itself, right next to the data, because that's where you have the raw data, that's where you have the maximum granularity of data, that's where you need to do the processing of which faces are which, right close to the edge itself, and then you can send the other data back up to the cloud, for example, to improve those algorithms within that camera, to do all that sort of work on the batch basis over time, that's what I was looking at, and looking at the cost justification for doing that sort of work. So today, we've got a set people here on the panel, and we want to talk about coming down one level to where IoT and IT are going to have to connect together. So on the panel I've got, I'm going to get these names really wrong, Sanjeev Kumar? >> Yes, that's right. >> From FogHorn, could you introduce yourself and what you're doing where the data is meeting the people and the machines? >> Sure, sure, so my name is Sanjeev Kumar, I actually run engineering for a company called FogHorn Systems, we are actually bringing analytics and machine learning to the edge, and, so our goal and motto is to take computing to where the data is, than the other way around. So it's a two-year-old company that started, was incubated in the hive, and we are in the process of getting our second release of the product out shortly. >> Excellent, so let me start at the other end, Rohan, can you talk about your company and what contribution you're focusing on? >> Sure, I'm head product marketing for Maana, Maana is a startup, about three years old, what we're doing is we're offering an enterprise platform for large enterprises, we're helping the likes of Shell and Maersk and Chevron digitally transform, and that simply means putting the focus on subject matter experts, putting the focus on the people, and data's definitely an important part of it, but allowing them to bring their expertise into the decision flows, so that ultimately the key decisions that are driving the revenue for these behemoths, are made at a higher quality and faster. >> Excellent. Well, two software companies, we have a practitioner here who is actually doing fog computing, doing it for real, has been doing it for some time, so could you like, Janet George from Western Digital, can you introduce yourself, and say something from the trenches, of what's really going on? >> Okay, very good, thank you. I actually build infrastructure for the edge to deal with fog computing, and so for Western Digital, we're very lucky, because we are the largest storage manufacture, and we have what we call Internet of Things, and Internet of Test Equipment, and I process petabytes of data that comes out of the Internet of Things, which is basically our factories, and then I take these petabytes of data, I process them both on the cloud and then on the edge, but primarily, to be able to consume that data. And the way we consume that data is by building very high-profile models through artificial intelligence and machine learning, and I'll talk a lot more about that, but at the end of the day, it's all about consuming the data that you collect from anywhere, Internet of Things, computer equipment, data that's being produced through products, you have to figure out a way to compute that, and the cloud has many advantages and many trade-offs, and so we're going to talk about the trade-offs, that's where the gap for computing comes into play. >> Excellent, thanks very much. And last but not least, we have Val, and I can never pronounce your surname. >> Bercovici. >> Thank you. (chuckling) You are in the midst of a transition yourself, so talk about where you have been and where you're going. >> For the better part of this century, I've been with NetApp, working at various functions, obviously enterprise storage, and around 2008, my developer instinct kind of fired up, and this thing called cloud became very interesting to me. So I became a self-anointed cloud czar at NetApp, and I ended up initiating a lot of our projects which we know today as the NetApp Data Fabric, that culminated about 18 months ago, in acquisition of SolidFire, and I'm now the acting CTO of SolidFire, but I plan to retire from the storage industry at the end of our fiscal year, at the end of April, and I'm spending a lot of time with particularly the Cloud Native Compute Foundation, that is, the opensource home of Google's Kubernetes Technology and about seven other related projects, we keep adding some almost every month, and I'm starting to lose track, and spending a lot of time on the data gravity challenge. It's a challenge in the cloud, it's a particularly new and interesting challenge at the edge, and I look forward to talking about that. >> Okay, and data gravity is absolutely key, isn't it, it's extremely expensive and extremely heavy to move around. >> And the best analogy is workloads are like electricity, they move fairly easily and lightly, data's like water, it's really hard to move, particularly large bodies around. >> Great. I want to start with one question though, just in the problem, the core problem, particularly in established industries, of how do we get change to work? In an IT shop, we have enough problems dealing with operations and development. In the industrial world, we have the IT and the OT, who look at each other with less than pleasure, and mainly disdain. How do we solve the people problem in trying to put together solutions? You must be right in the middle of it, would you like to start with that question? >> Absolutely, so we are 26 years old, probably more than that, but we have very old and new mix of manufacturing equipment, it's a storage industry, and in our storage industry, we are used to doing things a certain way. We have existing data, we have historical data, we have trend data, you can't get rid of what you already have. The goal is to make connectors such that you can move from where you're at to where you're going, and so you have to be able to take care of the shift that is happening in the market, so at the end of the day, if you look at five years from now, it's all going to be machine learning and AI, right? Agent technology's already here, it's proven, we can see, Siri is out here, we can see Alexa, we can see these agent technologies out there, so machine learning is a getting a lot of momentum, deep learning and neural networks, things like that. So we got to be able to look at that data and tap into our data, near realistically, very different, and the way to do that is really making these connections happen, tapping into old versus new. Like for example, if you look at storage, you have file storage, you have block storage, and then you have object storage, right? We've not really tapped into the field of object storage, and the reason is because if you are going to process one trillion objects like Amazon is doing right now with S3, you can't do it with the file system level storage or with the blog system level storage, you have to go to objects. Think Internet of Things. How many trillions of objects are going to come out of these Internet of Things? So one, you have to be positioned from an infrastructure standpoint. Two, you have to be positioned from a use case prototyping perspective, and three, you got to be able to scale that very rapidly, very quickly, and that's how change happens, change does not happen because you ask somebody to change their behavior, change happens when you show value, and people are so eager to get that value out of what you've shown them in real life, that they are so quick to adapt. >> That's an excellent-- >> If I could comment on that as well, which is, we just got through training a bunch of OT guys on our software, and two analogies that actually work very well, one is sort of, the operational people are very familiar with circuit diagrams, and so, and sort of, flow of things through essentially black boxes, you can think of these as something that has a bunch of inputs and has a bunch of outputs. So that's one thing that worked very well. The second thing that works very well is the PLC model, and there are direct analogies between PLC's and analytics, which people on the floor can actually relate to. So if you have software that's basically based on data streams and time, as a first-class citizen, the PLC model again works very well in terms of explaining the new software to the OT people. >> Excellent, okay, would you want to come in on that as well? >> Sure, I think a couple of points to add to what Janet said, I couldn't agree more in terms of the result, I think Maana did a few projects, a few pilots to convince customers of their value, and we typically focus very heavily on operationalizing the output, so we are very focused on making sure that there is some measurable value that comes out of it, and it's not until the end user started seeing that value that they were willing and open to adopt the newer methodologies. A second point to that is, a lot of the more recent techniques available to solve certain challenges, there are deep learning neural nets there's all sorts of sophisticated AI and machine learning algorithms that are out there, a lot of these are very sophisticated in their ability to deliver results, but not necessarily in the transparency of how you got that, and I think that's another thing that Maana's learning, is yes, we have this arsenal of fantastic algorithms to throw at problems, but we try to start with the simplest approach first, we don't unnecessarily try to brute force, because I think an enterprise, they are more than willing to have that transparency in how they're solving something, so if they're able to see how they were able to get to us, how the software was able to get to a certain conclusion, then they are a lot happier with that approach. >> Could you maybe just give one example, a real-world example, make it a little bit real? >> Right, absolutely, so we did a project for a very large organization for collections, they have a lot of outstanding capital locked up and customers not paying, it's a standard problem, you're going to find it in pretty much any industry, and so for that outstanding invoice, what we did was we went ahead and we worked with the subject matter experts, we looked at all the historical accounts receivable data, we took data from a lot of other sources, and we were able to come up with models to predict when certain customers are likely to pay, and when they should be contacted. Ultimately, what we wanted to give the collection agent were a list of customers to call. It was fairly straightforward, of course, the solution was not very, very easy, but at least on a holistic level, it made a lot of sense to us. When we went to the collection agents, many of them actually refused to use that approach, and this is part of change management in some sense, they were so used to doing things their way, they were so used to trying to target the customers with the largest outstanding invoice, or the ones that hadn't paid for the longest amount of time, that it actually took us a while, because initially, what the feedback we got was that your approach is not working, we're not seeing the results. And when we dug into it, it was because it wasn't being used, so that would be one example. >> So again, proof points that you will actually get results from this. >> Absolutely, and the transparency, I think we actually sent some of our engineers to work with the collections agents to help them understand what approach is it that we're taking, and we showed them that this is not magic, we're actually, instead of looking at the final dollar value, we're looking, we're calculating time value lost, so we are coming up with a metric that allows us to incorporate not just the outstanding amount, or the time that they haven't paid for, but a lot of other factors as well. >> Excellent, Val. >> When you asked that question, I immediately went to more of a nontechnical business side of my brain to answer it, so my experience over the years has been particularly during major industry transitions, I'm old enough to remember the mainframe to client server transition, and now client server to virtualization and cloud, and really, sales reps have that well-earned reputation of being coin-operated, though it's remarkable how much you can adjust compensation plans for pretty much anyone, in a capitalist environment, and the IT/OT divide, if you will, is pretty easy to solve from a business perspective when you take someone with an IT supporting the business mentality, and you compensate them on new revenue streams, new business, all of a sudden, the world perspective changes sometimes overnight, or certainly when that contract is signed. That's probably the number one thing you can do from a people perspective, is incent them and motivate them to focus on these new things, the technology is, particularly nowadays is evolving to support them for these new initiatives, but nothing motivates like the right compensation plan. >> Excellent, a great series of different viewpoints. So the second question I have again coming down a bit to this level, is how do we architect a solution? We heard you got to architect it, and you've got less, like this, it seems to me that that's pretty difficult to do ahead of where you're going, that in general, you take smaller steps, one step at a time, you solve one problem, you go on to the next. Am I right in that? If I am, how would you suggest the people go about this decision-making of putting architectures together, and if you think I'm wrong and you have a great new way of doing it, I'd love to hear about it. >> I can take a shorter route. So we have a number of customers that are trying to adopt, are going through a phased way of adopting our technology and products, and so it begins with first gathering of the data, and replaying it back, to build the first level of confidence, in the sense that the product is actually doing what you're expecting it to do. So that's more from monitoring administration standpoint. The second stage is you should begin to capture analytical logic into the project, where it can start doing prediction for you, so you go into, so from operational, you go into a predictive maintenance, predictive maintenance, predictive models standpoint. The third part is prescriptive, where you actually help create a machine learning model, now, it's still in flux in terms of where the model gets created, whether it's on the cloud, in a central fashion, or some sort of a, the right place, the right context in a multi-level hierarchical fog layer, and then, you sort of operationalize that as close to the data again as possible, so you go through this operational to predictive to prescriptive adoption of the technology, and that's how people actually build confidence in terms of adopting something new into, let's say, a manufacturing environment, or things that are pretty expensive, so I give you another example where you have the case of capacitors being built on a assembly line, manufacturing, and so how do you, can you look at data across different stations and manufacturing on a assembly line? And can you predict on the second station that it's going to fail on the eighth one? By that, what you're doing is you are actually reducing the scrap that's coming off of the assembly line. So, that's the kind of usage that you're going to in the second and third stage. >> Host: Excellent. Janet, do you want to go on? >> Yeah, I agree and I have a slightly different point of view also. I think architecture's very difficult, it's like Thomas Edison, he spent a lot of time creating negative knowledge to get to that positive knowledge, and so that's kind of the way it is in the trenches, we spend a lot of time trying to think through, the keyword that comes to mind is abstraction layers, because where we came from, everything was tightly coupled, and tightly coupled, computer and storage are tightly coupled, structured and unstructured data are tightly coupled, they're tightly coupled with the database, schema is tightly coupled, so now we are going into this world of everything being decoupled. In that, multiple things, multiple operating systems should be able to use your storage. Multiple models should be able to use your data. You cannot structure your data in any kind of way that is customized to one particular model. Many models have to run on that data on the fly, retrain itself, and then run again, so when you think about that, you think about what suits best to stay in the cloud, maybe large amounts of training data, schema that's already processed can stay on the cloud. Schema that is very dynamic, schema that is on the fly, that you need to read, and data that's coming at you from the Internet of Things that's changing, I call it heteroscedastic data, which is very statistical in nature, and highly variable in nature, you don't have time to sit there and create rows and columns and structure this data and put it into some sort of a structured set, you need to have a data lake, you need to have a stack on top of that data lake that can then adapt, create metadata, process that data and make it available for your models, so, and then over time, like I totally believe that now we're running into near realtime compute bottleneck, processing all this pattern processing for the different models and training sets, so we need a stack that we can quickly replace with GPUs, which is where the future is going, with pattern processing and machine learning, so your architecture has to be extremely flexible, high layers of abstraction, ability to train and grow and iterate. >> Excellent. Do you want to go next? >> So I'll be a broken record, back to data gravity, I think in an edge context, you really got to look at the cost of processing data is orders of magnitude less than moving it or even storing it, and so I think that the real urgency, I don't know, there's 90% that think of data at the edge is kind of wasted, you can filter through it and find that signal through the noise, so processing data to make sure that you're dealing with really good data at the edge first, figuring out what's worth retaining for future steps, I love the manufacturing example, I have lots of customer examples ourselves where, for quality control in a high-moving assembly line, you want to take thousands of not millions of images and compare frame and frame exactly according to the schematics where the device is compared to where it should be, or where the components, and the device compared to where they should be, processing all of that data locally and making sure you extract the maximum value before you move data to a central data lake to correlate it against other anomalies or other similarities, that's really key, so really focus on that cost of moving and storing data, yeah. >> Yes, do you want the last word? >> Sure, Maana takes an interesting approach, I'm going to up-level a little bit. Whenever we are faced with a customer or a particular problem for a customer, we try to go over the question-answer approach, so we start with taking a very specific business question, we don't look at what data sources are available, we don't ask them whether they have a data lake, or we literally get their business leaders, their subject matter experts, we literally lock them up in a room and we say, "You have to define "a very specific problem statement "from which we start working backwards," each problem statement can be then broken down into questions, and what we believe is any question can be answered by a series of models, you talked about models, we go beyond just data models, we believe anything in the real world, in the case of, let's say, manufacturing, since we're talking about it, any smallest component of a machine should be represented in the form of a concept, relationships between people operating that machinery should be represented in the form of models, and even physics equations that are going into predicting behavior should be able to represent in the form of a model, so ultimately, what that allows us is that granularity, that abstraction that you were talking about, that it shouldn't matter what the data source is, any model should be able to plug into any data source, or any more sophisticated bigger model, I'll give you an example of that, we started solving a problem of predictive maintenance for a very large customer, and while we were solving that predictive maintenance problem, we came up with a number of models to go ahead and solve that problem. We soon realized that within that enterprise, there are several related problems, for example, replacement of part inventory management, so now that you figured out which machine is going to fail at roughly what instance of time from now, we can also figure out what parts are likely to fail, so now you don't have to go ahead and order a ton of replacement parts, because you know what parts are going to likely fail, and then you can take that a step further by figuring out which equipment engineer has the skillset to go ahead and solve that particular issue. Now, all of that, in today's world, is somewhat happening in some companies, but it is actually a series of point solutions that are not talking to each other, that's where our pattern technology graph is coming into play where each and every model is actually a note on the graph including computational models, so once you build 10 models to solve that first problem, you can reuse some of them to solve the second and third, so it's a time-to-value advantage. >> Well, you've been a fantastic panel, I think these guys would like to get to a drink at the bar, and there's an opportunity to talk to you people, I think this conversation could go on for a long, long time, there's so much to learn and so much to share in this particular information. So with that, over to you! >> I'll just wrap it up real quick, thanks everyone, give the panel a hand, great job. Thanks for coming out, we have drinks for the next hour or two here, so feel free to network and mingle, great questions to ask them privately one-on-one, or just have a great conversation, and thanks for coming, we really appreciate it, for our Big Data SV Event livestreamed out, it'll be on demand on YouTube.com/siliconangle, all the video, if you want to go back, look at the presentations, go to YouTube.com/siliconangle, and of course, siliconangle.com, and Wikibond.com for the research and content coverage, so thanks for coming, one more time, big round of applause for the panel, enjoy your evening, thanks so much.