>> Announcer: From around the globe, it's theCUBE. With digital coverage of Dockercon live 2020. Brought to you by, Docker, and its ecosystem partners. >> Hey, welcome back everyone to the DockerCon 2020, #DockerCon20. This is The Cube virtual coverage with Docker on their event here. And we're in the studio in Palo Alto, I'm John Furrier your host of theCUBE, we're here with a great guest to talk about Docker Desktop, the Microsoft relationship, and the key news that's coming out. Ben De St Paer-Gotch is the product manager for Docker Desktop. Ben, great for coming on, thanks for spending the time with me. >> Thanks for having me, I really appreciate it. >> So obviously, this is a virtual conference, we wish we could be in person, but given the state of affairs we're going to do remotely, but the momentum Docker has is phenomenal, it's always been great with containers. It's the number one downloaded app around for developers. Microsoft just had their Build conference, which was again virtual as well, or digital, as they say, it's interchangeable. But clear momentum now with Docker as containers actually is the standard, you guys are doing great. What's the key news out of the Microsoft world for people who missed it last week with MS Build? >> Yeah, so last year at Build, Microsoft announced WSO2 to the Windows subsystem with Linux two. (mumbles) The mapping between the windows (mumbles) Which, went really well but it just didn't provide the same centered needed Linux experience. Last year, they announced Windows subsystem Linux two, (Provides an actual Linux one on windows machine, and we've been working hard with Microsoft over the last year to integrate proper desktop as a main desktop application for working with containers with WSO2. A build this year, Microsoft has gone on and announced that WSO2 is going to have a few new features, and it's going to have new features. (mumbles) Mention Linux graphical, Linux applications, you can access the file system, the installation is going to become a slicker which I guess I'm the most excited about that pitch. But the most exciting announcement is, they will be bringing GPU support to WSO2 which means that we will be able to provide and give you support through Docker desktop or container workloads that peoples are working on. And now we're launching Gray and Agua through containers and docks and desktops and Windows which is really cool because we haven't been able to do that before. >> So is this the first GPU support on Microsoft Windows for Docker, with Docker? >> It's, yeah, it's the first GPU Support for Docker Desktop or Mac or Windows. So, previously the hypervisor hasn't passed through the GPU, pretty much, which meant that we couldn't access it from Docker desktop. So Docker desktop isn't about a lightweight VM we sorts of plumb all that in for you. But we're limited about what we could get access to from the hypervisor, Microsoft putting this through and giving us access for the first time, we can actually, we can go. >> Not to go on a side tangent here, but you know, all these virtual events, and I was watching some of the build stuff as well, as well as us immediate streamers and doing stuff, you can see people's home rigs. And you talk to any Developer, video streamer, or anyone who is working remotely, if you don't have the best GPU's in there, I mean, this has just become, I mean, quite frankly, you need the GPU's. So this is important, it's not only from a vanity standpoint performance. Having that support, I'm going to want the best GPU's, I'm always going to be upgrading my machine for that extra power. What's the impact? What does it mean for me as a Developer? Does it increase stuff? What's the bottom line? >> As a Developer, it means you actually have access to it. So, especially when you're doing workloads on the CPU, you've got minimum amounts of power utilization you can do. When you're running workloads for an L Development, you have a lot of power up process you've got to log, to do your mobile training. So, in an element cycle, you're likely to have your application which you're going to use to produce a modeling, you're going to have training data. Taking that training data and producing a model requires lots of panel processing which is an enormous calculations in producing with finer waitings. Doing that on a CPU has to be done on a serial fashion rather than parallel, which is huge and intensive and takes a really long time. Whereas on a GPU, you can do all of that in parallel which massively reduces the amount of time it will take to run those training functions. Either just straight up in Linux or running them in a container, which as the more of people are looking at running container with workloads, it's how I first, the first team that I was on actually used Docker. I was working in Amazon Alexa, and my team picked up the opportunity to run our workload in container. And that was my first experience, so even though my team backed down, so I could see the system. >> Yeah, ML workloads automations could be critical of that performance. Okay, let's get into some of the momentum with Microsoft, you guys have obviously, builds over, we're here now at DockerCon, there's news. Could you share some of the tidbits for what's being talked about now with Docker and DockerCon. >> Yeah, absolutely, so, along with everything else we've been doing, we've been partnering with Microsoft trying to make the best experience generally with Docker desktop, and with WSO2 and with the VSCO. I've been working closely with Microsoft guys to actually try and improve our experience in Windows as it is today, and to improve some of those integrations with VSCO, and also working with the VSCO team on the Docker plugin for VSCO to give our feedback, and to hear feedback from those guys on the errors and issues they're seeing with Docker desktop and to really try to produce the best experience we can on Windows. End to end, from very front end running all the way through that first push, that first run on the cloud using Docker. >> So what is some of the new product management processes and customer support things that you guys are doing? This comes up a lot, obviously, we had a great conversation around shift left with security. That's great news there. You start to see a lot of this added value for Developers, wanted their support right? So how do I get things I need, and from a customer standpoint? It's kind of a moving train this world and it's only getting better and better from a Developer standpoint. But there's more complexity, it's got to be abstract the way you've got, you know, this new abstraction layers developing. You've got a lot of automation. How does the customer get the support they need in the same agile way that Developers are cranking out code? >> It's a really good question, it's something I think we're still working on as well. So, we're trying to working out and one of the big things I'm trying to work out is, how to make it easier for people to get started with Docker, and how do we also make sure with the things we build, we don't leave a cliff edge instead of a lining path. You don't get to a certain point in an easy process, and then the next step, takes you straight off a cliff, so that's not useful for anyone. So, producing those parts and those ways for people to learn and actually progress is something we're really trying to work out. How to make it natural from the first experience all the way through. From an actual support perspective, the other thing we're looking at, is we're trying to do more things in the open. We're really trying at Docker to bring as many of the new features and pieces we're developing which we have to do that in the open with community visibility, so that if people really want it fixed, they can open the PR and they can help us out. And then the last thing that my team really stood out was our Docker of having actions. As creators, someone already finished, could you do this? Someone else had a PR and emerged it. So, to a certain extent, you've got your one side which had you on board and this ever growth spiral and you keep learning. The other side is how'd you fix the board when you find an issue? In that one, we're really trying to work with the community, a lot more than we have in the last couple of years. >> Awesome, some folks watching, hit him up on Twitter, he's the Product Manager for Docker Desktop among other things. You guys are very transparent, you've got your Twitter handle on the lower third. People can chime in or just jump on the chat, we'll follow up and get you the info. Final question for you Ben, as you look at this reality we're in, there's kind of a holistic kind of moment now where people kind of realizing the new realities here. You're looking at the.. you get the keys to the kingdom with Docker Desktop, okay. You got some momentum with Microsoft, the developer role is moving fast and fast as the head room increases for capabilities with automation. And I know you mentioned a few of those things. GPU is now available. What's the future look like for these Developers? The next short, medium and long term? What's your view as you look out over the landscape because you've got to look at the product roadmap, your engagement with the community. Can you share some insight into how you're thinking about Docker Desktop going forward? >> Yeah, absolutely. So, I think what really interesting point as you say, which is that, if you look at sort of a lot of the Developer side of things that have sort of come out in the last like six months, six to eighteen months. The things I see, I see daily like you mention, things like orchestrating for containers gaining momentum. If you think about crossing the Kaizen model, we're just passed the early Dockers now. We're kind of into the early majority, but we're going to start to move over the next few years into the late majority. What that really means is that people here have been using one of two of these technologies. Maybe you've been using cloud, maybe you've been using Edge, maybe you've been using containers, maybe you've been using CICD, maybe you are using Expiration, maybe you're not. Maybe you've got a Microservice application, maybe it's a little bit of a mole rat. What we're really going to see is, you're going to start to see, all of these changes intersecting and overlapping. And people who have started to pick up model two of these will start to pick up all of them. And that's probably going to happen as we move into the majority of users. So from a what's coming instead of a lot of those thing that you see in best practice in the ideal Developer setup, so a beautiful CICD, a more of an orchestrated environment, Microservice architecture, we're going to see a lot more of that becoming the norm. But I think along with that, we'll also see a level of recognition coming along that a single Microservice alone doesn't provide value. And that's it's going to be some of those groups of services that will provide the user outcome. And that's where my focus is at the end which is you know, an authentication service is great but it doesn't provide value unless you give access to something as authentic. >> It's been issued that the new Docker is all about Developer experience. This is really the core mission. I mean, since the sale of the piece of morantis, Docker has retrenched and reinvented, but stayed core to its principles. Just share with the Developers who've been watching that are coming back into the ecosystem, what is this new Docker vibe? Share your thoughts. >> The new Docker vibe is about working in the open, and it's about solving problems for Developments. The original goal of Docker was to make it easy to pack and ship. It was to reduce Developer friction. As we move more into, sort of, the enterprise space, we worry more about Ops and DevOps. We're not trying to re-focus on Developer and if you sort of think there's two parts to the Developer life cycle, where you've got your work, where you're doing your creative work, where you're writing code. And then you've sort of got your part of the inner loop. And then you've got your part where you're trying to get that code out to production, you're trying to get your value to someone else. Instead of your outer loop, we're really trying to focus on the inner loop And sort of our mantra is that any bit for a Developer should spend as much as their time as possible creating new and exciting things and we're onto those holes that reduce those boring, Monday, repetitive tasks, that we're really trying to work out how we take those boring repetitive pieces and how do we make them just vanish like magic from new users or how do we reduce the friction for the experience from users? From both desktop and hub, we're really trying to bring those two together to achieve that. >> You know what's great about folks who have been in the class since day one. All of us have scar tissue experiences, you know the one thing that's constant is constant change. And one of the things that you guys have done at Docker, and hats off to the whole, you know, original team, is that brand of Docker has symbolized quality openness, and set the standard, I mean, if you look back and containers were really coming around, it's not a new concept. But Docker really set the industry on this path and it's been great to follow every DockerCon at TheCube coverage, but more importantly, as the demand for Developers to build these next wave of Cambrian explosion of applications. It's going to be more important than ever to have more of these abstractions, more of these tools in this real time, more Developers experience because there's more building going on. And it's not just one cloud, it's all clouds, it's all things. >> Yeah, I think it was like when IDC analyzed the future report a couple years ago, I think it was maybe the 2018 one. They said that maybe 2017. They said to date, we've built 500 millions applications worldwide and by 2023, we'll build another 500 million. The rate of creation is just insane, it's exponential growth of us producing more and more applications and connecting more and more devices to do them. The sheer volume of creation and the rate of new technology supporting, even with the rate of companies adopting, I guess more of a warm cloud. I think it's like 60 percent of companies are now more than one cloud provider. Maybe even more, maybe it's like 80 percent. It's ridiculous. >> I was just having this debate on Twitter about this multi-cloud. Someone tried to call us out saying, "Oh you guys were pooing on multi-cloud in 2016 and 18." I go "Look at, no one was Pooping on multi-cloud, it didn't exist." I had multiple clouds but there was no real use case. Now you're starting to see the use cases, where yeah, I had multiple clouds and I got Azure here, I got this over here. But no one wakes up and spreads their workloads wrong. This is going back a few years. Certainly the hybrid was developing, but I think now you're starting to see with networking and some of these inter-operable dynamics, you start to see innovation pockets in wide spaces in large market opportunities for start-ups and companies to thread the clouds together at the right place. So I think multi-cloud is becoming apparent from a use case stand point. Still a ton of work to do, I mean direct connects, got SLA's, I mean all kinds of stuff at the networking level but it is real. It's going to be one of those realities that everyone has, at least one or two, if not three. It could be optimization, this is what Developers do right? Solve problems. >> Yeah, absolutely, I mean if nothing else, I've encounter a couple of companies even just where redundancy is handled by multi-cloud strategy. If you want to achieve more nines and you're just balancing workloads between two clouds. >> I mean, the Zoom news was really a testament to that because everyone got into a twist over that. Oh Zoom moves off Amazon, no they didn't move off Amazon, they went to Oracle, they got Adge, they're everywhere. Why wouldn't they be? They need to pass it, they fail over, they need fall tolerance, I mean, these are basic distributing computing concepts that is one on one. You've got to have these co-locations. And optimization for those clouds and the apps on Microsoft as well, so why wouldn't you do it? >> Exactly. And that's that hybrid, that multi-cloud, compounding that some of which you said earlier, that over changes when you're looking at how you go to CICD, how you're bundling these applications, creating more applications than ever. Coming back, sort of, with more AI workloads, much like GPU and you combine that with, sort of, last in the growth of age devices as well. It sort of makes for a really interesting future. And Docker is sort of, that summation SOV, what we're using to frame how we're thinking about our product and what we should be building. >> Great, for the audience out there, hit him up on Twitter, Ben's available, they're out in the open, if you're interested in how Docker makes life easier on the Windows platform, with the GPU support, they've got security now built in, shifting left. Give these guys a call and of course, we love the mission, out in the open. It's theCUBE's mission as well and great to chat with you. Ben, thanks for spending the time with me today. >> Been an absolute pleasure, thank you for having me. >> Okay, just TheCube's coverage, the virtual Cube with DockerCon co-creating together out in the open. DockerCon20, #Docker20, I'm John Fer with TheCube, stay tuned for our next segment, and thanks for watching. (ambient music)

>> live from Washington, D. C. It's the Cube covering a ws public sector summit by Amazon Web services. >> Welcome back, everyone. You are watching the Cuban. We're kicking off our day two of our live coverage a ws public sector summit here in our nation's capital. I'm Rebecca Knight co hosting with John Fer Yer John. It's great to be here. 18,000 people having important conversations around around governments and cloud computing. Let's extract the signal from the noise. Let's do with the Cube. Does best, >> Yeah, I mean, this is to me a really exciting event because it's got the confluence of what we love tech and cloud computing and all the awesomeness of that and that enables. But even in Washington, D. C. With the backdrop against tech clash on this, you know, narrative run tech for illah tech for bad, bad check whatever you want to call it. Anti trust is a lot of narratives around that there's a huge story around check for good. So I think there's an interesting balance there around the conversations, but this is world of heavy hitters are this week You've got senior people at the government level here, you have senior tech people hear all kind of meddling and trying to figure out howto let the tail winds of cloud computing Dr Change within government against this backdrop of tech for ill as Jay Carney, whose the global marketing policy guy for Amazon on reports to Jeff Bezos, former Obama press secretary. He's super savvy on policy, super savvy on tech. But this is a really big point in time where the future's gonna be determined by some key people and some key decisions around the role of technology for society, for the citizens, United States, for nation states as people start to figure out the role of data and all the impact of this so super exciting at that level, but also dangerous and people are telling a little bit. But I also want to run hard. That's pretty much the big story. >> So let's let's let's get into this tech backlash because you're absolutely right. Through the public, sentiment about technology and the tech behemoths has really soured. The regulators are sharpening their blades and really paying much more attention, uh, particularly because so many people say, Hey, wait a minute, why? How does Google and Facebook know all this stuff about me, but what do you think? What are we hearing on the ground in terms of where regulation is going? Before, before the cameras were rolling, you were talking about this idea of regulators working closely with the innovators, observing but not meddling. I mean, do you think that that's that's That's these dollars underwears We're going in? >> Well, not really. I think that that's where people wanted to go in. I think right now the the surprise attack of tech taking over, if you will in the minds of people and or without Israel or not, it's happened, right? So I was talking yesterday around how the Internet, when Bill Clinton was president, really grew a little bit slower than the pace of this today. But they did a good job of managing that they had private sectors take over the domain name system. We saw that grow that created in the open Web and the Web was open. Today it's different. It's faster in terms of technology innovation, and it's not as open. You have Facebook, LinkedIn and these companies that have silos of data, and they're not sharing it with cyber security General Keith Alexander, former head of the NSA and the first commander of cyber command in the U. S. The United States under Obama. He pointed out that visibility into the cyber attacks aren't there because there's no sharing of data. We heard about open data and knishes from a think tank. The role of data and information is going to be a critical conversation, and I don't think the government officials are smart enough and educated enough yet to understand that So regulatory groups want to regulate they don't know how to. They're reaching out the Amazons, Google's and the Facebook to try to figure out what's going on. And then from there they might get a path. But they're still in the early stages. Amazon feels like they're not harming anyone there. Lower prices, fast delivery, more options. They're creating an enablement environment for tons of startups, so they feel like they're not harming anyone. You're the antitrust, but if they're going to being monopolizing the market place, that's another issue. But I still think Amazon still an enabling mode, and I think you know, they're just running so hard. It's going so fast, I think there's gonna be a big challenge. And if industry doesn't step up and partner with government, it's going to be a real mess. And I think it's just moving too fast. It's very complicated. Digital is nuanced. Now. You get the role of data all this place into into into effect there. >> Well, you're absolutely right that it's going fast. Teresa Carlson on the other day talking about eight of US growth, UH, 41% year over year and she said, Cloud is the new normal. The cloud cloud is here more and more governments on state and local, really recognizing and obviously international countries to recognize that this, this is they're adopting these cloud first approach is, >> yeah, I mean, I think the first approach is validated 100%. There's no debate. I think it's not an ah ha moment. Cloud Israel. Amazon has absolutely proven since the CIA deal in 2013 that this is a viable strategy for government to get to value fast, and that is the whole speed of cloud game. It's all about time to value with agility. Eccentric center. We've been talking about that with Dev Ops for a long, long time. The real thing that I think's happening that's going on. That's kind of, you know, to read the tea leaves and we'll hear from Corey Quinn. Our host at large will go on later. This is a new generation of talent coming on board and this new generation. It feels like a counterculture mindset. These are Dev ops, mindset, people not necessarily Dev ops like in the Cloud Computing Way. They're younger, they're thinking differently, and they think like Amazon not because they love Amazon, because that's their nature. Their got their getting content in a digital way, their digital natives. They're born into that kind of cultural mindset. Of what is all this nonsense red tape? What's the bottlenecked in solving these problems? There's really not a good answer anymore, because with cloud computing and machine learning an A I, you can solve things faster. So if you expose the data, smart people go well. That's a problem that could be song. Let's solve it. So I think there's going to be a resurgence is going to be a renaissance of of younger people, kind of in a counter culture way that's going to move fast and an impact society and I think it's gonna happen pretty quickly over the next 10 years. >> Well, that's one of the things that's so inspiring about being at a conference like this one a ws public sector summit, Because we are hearing getting back to what you just said. We're solving problems and these air problems about not just selling more widgets. This's actually about saving lives, helping people, delivery of healthcare, finding Mr Missing Persons and POWs who are missing in action. >> I mean, the problems could be solved with technology now for goodwill, I think will outweigh the technology for Ayla's Jay Carney calls it. So right now, unfortunately, was talking about Facebook and all this nonsense that happened with the elections. I think that's pretty visible. That's painful for people to kind of deal with. But in the reality that never should have happened, I think you're going to see a resurgence of people that's going to solve problems. And if you look at the software developer persona over the past 10 to 15 years, it went from hire. Some developers build a product ship it market. It makes some money to developers being the frontlines. Power players in software companies there on the front lines. They're making changes. They're moving fast, creating value. I see that kind of paradigm hitting normal people where they can impact change like a developer would foran application in society. I think you're gonna have younger people solving all kinds of crisis around. Whether it's open opioid crisis, healthcare, these problems will be solved. I think cloud computing with a I and machine learning and the role of data will be a big catalyst. >> But money, the money, the money is the thing we're going to have Cory Quinn on later talking about this this talent gap because there are people who are, As you said, they're young people who are motivated to solve these problems, and they want to work for mission driving institutions. What better mission, then helping the United States government >> just heard in the hallway? This has been the I've heard this multiple times here. This show I just heard someone saying Yeah, but that person's great. I can't keep them. What's happening is with the talent is the people that they need for cloud computing. Khun, get a job that pays three times Mohr orm or at the private sector. So, you know, Governor doesn't have stock options, >> right? All right, all right. If >> you're, ah, machine learning, >> people call girls in the lounge. >> Eso all kinds of different diners. But I think this mission driven culture of working for society for good might be that currency. That will be the equivalent stock option that I think is something that we were watching. Not haven't seen anything yet, But maybe that will happen. >> Paid in good feelings way. We've got a lot of great guests. Wave already teed up. We've got your E. Quinn. Bill Britain from Cal Poly to talk more about ground station. We have alien Gemma Smith of YSL Itics, uh, and Jameel Jaffer. >> Think ground station. But the biggest surprise for me and the show so far has been ground station that that product has got so much traction. That's ridiculous. I thought it would be kind of cool. Spacey. I like it, but it's turning into a critical need for a I ot I mean, I was just talking with you. Came on about the airplane having WiFi on the plane. We all like Wow, we expected now, but you go back years ago is like, Oh, my God. I got WiFi on the plane. That's a ground station, like dynamic people going. Oh, my God. I can provision satellite and get data back, all for io ti anywhere in the world. So that is pretty killer. >> Excellently. I'm looking forward to digging in with you with many guests today. >> Good. >> I'm Rebecca Knight. For John. For your stay tuned, you are watching the Cube.

I'm from Seattle Washington it's the cube covering Gube Khan and cloud native Khan North America 2018 brought to you by Red Hat the cloud native computing foundation and its ecosystem partners ok welcome back everyone we are here live with cube covers three days with wall-to-wall coverage here at coop con cloud native con 2018 in Seattle I'm John fer with the cubes to Minutemen here breaking it down we're at day two we've got a lot of action David Ronn chick who's the head of open source ml strategy at Azure at Microsoft Microsoft Azure formerly of Google now at Microsoft welcome back to the cube we had a great chat at Copenhagen good to see you great to see you too thank you so much for having me you've been there from day one it's still kind of day one in Korea is still growing you got a new gig here at Microsoft formerly at Google you had a great talk at Google next by the way which we watched and and caught on online you just you're still doing the same thing think of me to explain kind of what the new job is what your focus is absolutely so in many ways I'm doing a very similar job to the one I was doing at Google except now across all of Asher you know when you look at machine learning today the truth of the matter is is it is about open source it's about pulling in the best from academia and open source contributors developers across the spectrum and while I was at Google I was able to launch the cube flow project which solves the very specific but very important problem now that you look at Azure a company that is growing excuse me a division that is growing extremely quickly and looking to expand their overall open source offerings make investments work with partners and projects and make sure that that researchers and customers are able to get to machine learning solutions very quickly I'm coming in to help them think about how to make those investments and accelerate customers overall time to solutions so both on the commercial side Asscher which is got a business objective to make money but also open source how is it still open source for you is it all open sores or is it crossing a little bit of bulk just quickly clarify that yeah there's no question um you know obviously as you as a business they pay me a salary and and we're gonna have a great first party solution for all of these very things but the reality is much like kubernetes has both a commercial offering and an open-source offering I think that all the major cloud providers will have that kind of duality they'll work in open source and and you can measure you know how many contributions and what they're doing in the open source projects but then they'll also have hosted and other versions that make it easier for customers to migrate their data and adopt some of these new so you know one of the things that's interesting on that point is this a super important point is that open source community that's here with kubernetes around kubernetes it's all kind of upstream kind of concept but the downstream impacts our IT and your classic developer so you have your open source yeah and a thing going on that's the core of this community an event the IT investments are shifting in 2019 we are seeing the trend of somewhat radical but certainly a reimagining of the IT I mean certainly you guys have gone cloud at Azure has seen that that result absolutely good pick up by customers office 365 that's now a SAS that's now now you've got cloud you have cloud scale this is what machine learning is really shining so I the question to you is what do you think is gonna be the big impact of 2019 to IT investment strategies in terms of what they how they procure and consume technology how they build their apps with the new goodness coming in from kubernetes etc absolutely um you know I remember back in the day you know I was an IT admin myself and and I carried a pager for literally when you know a machine went down or a power supply went out or this Ram was bad or something like that today if you went to even the most sophisticated IT shop they would be like what are you crazy you you should never carry a pager for that you should have a system that understands it's ok if something that low-level goes out that's exactly what kubernetes provided it provided this abstraction layer on top of this so if you went down kubernetes knew had a reschedule a pod and move things back and forth taking that one step further now into machine learning unfortunately today people are carrying pagers for the equivalent of if a power supply goes out or something goes wrong it's still way too low-level we're asking data scientists ml engineers to think about how to provision pods how'd it work on drivers how to do all these very very low-level things with things like kubernetes with things like hume flow you're now able to give higher level abstraction so a data scientist can in and you know open up their Jupiter notebook work on the model see how it works and when they're done they hit a button and it will provision out all the machines necessary all the drivers all the everything spin it up run that training job and bring it back and shut everything down so they won't wonder if you can help expand on that a little bit more so you know what one of the things that that's great about kubernetes is it can live in a diverse amount of infrastructure one of the biggest challenges with machine learning is you know where's my data how do I get to the right place where do I do the training you know we've spending a lot a couple of years looking at you know edge and you know what's the connectivity and how we're gonna do this you help just kind of pan us picture the landscape and what do we have solved and what are we working at trying to get put together yeah you know I think that's a really excellent question today there's so much focus on well are you gonna choose pi torch or tensorflow CNT k MX net you know numpy scikit-learn there are a bunch of really great frameworks out there done in the open source and we're really excited but the reality is when you look at the overall landscape that's just 5% of the work that the average data scientist goes through exactly your point how do I get my data in how do I transform it how do I visualize it generate statistics on it make sure that it's not biased towards certain populations and then once I'm done training how do I roll it out to production and monitor it and log and all these things and that's really what we're talking about that's what we tried to get work on when it comes to cute flow is is to think about this in a much broader sense and so you take things like data the reality is you can't beat the speed of light if I have a petabyte of data here it's gonna take a long time to move it over there and so you're gonna be really thoughtful about those kind of things i I'm very hopeful that academic research and and industry will figure out ways to reduce the amount of data and make it much much more sane in overall addressing this problem and make it easier to train in various locations but the reality is is I think you're ultimately gonna have models and training and inference move to many many different locations and so you'll do inference at the edge on my phone or on a you know little Bluetooth device in the corner of my house saying whether or not it's too hot or too cold we're gonna need that kind of intelligence and we're gonna do that kind of training and data collection at the edge do you see a landscape evolving where you have specialty ml for instance like the big caution in IOT is move you know compute to the data yeah reads that latency you see machine learning models moving around at code so I can throw a machine learning at a problem and there's that and that is that what kubernetes fits and I'm trying to put together a mental model of how to think about how ml scales yeah what's your vision on that how do you see that evolving yeah absolutely I think that you know going back to what we talked about at the beginning we're really moving to much more of a solution driven architecture today ml you know is great and the academic research is phenomenal but it is academic research it didn't really start to take off until people invented things are you know creating things like image Nets and mobile net and things like that that did very important things like object detection but then people that you know commercial researchers were able to take that and move that into locations where people actually need it in I think you will continue to see that that migration I don't think you're gonna have single ml models that do a hundred different things you're gonna have a single ml model that does a vertical specific thing anomaly detection in whatever factories and you're gonna use that in a whole variety of locations rather than trying to you know develop 1 ml model to solve them all so it's application specific or vertical alright so that means the data is super important quality data clean data is clean results dirty date bad result absolutely right people have been in this kind of virtuous circle of cleaning data you know you guys know at Google certainly Microsoft as well you know datum data quality is critical but you got the horizontally scalable cloud but you need specialism around the data and for them ml how do you see that is that I mean obviously sounds like the right architecture this is where the finesse is and the nuance I don't see that so you know you you bring up a really interesting point today the the biggest problem is is how much data there is right it's not a matter of whether or not you're able to process it you are but but it's so easy to get lost caught and little anomalies you know if you have a petabyte of data and whatever a megabyte of it is the thing that's causing your model to go sideways that's really hard to detect I think what you're seeing right now is a lot of academic research which I'm very optimistic about that will ultimately reduce that that will both call out hey this particular data is smells kind of weird maybe take a closer look at this or you will see a smaller need for training you know where it was once a petabyte you're able to train on just 10 gigabytes I'm very optimistic that both of those things happen and as you start to get to that you get better signal-to-noise and you start saying oh in fact this is questionable data let's move that off to the side or spend more time on it rather than what happens today which is oh I got this model and it works pretty well I'm just going to throw everything at it and trying you know get some answer out and then we'll go from there and that's with a lot of false positives come in all absolutely all right so take the next level here at Kubb con cloud native con in this community where kubernetes is the center of all these sets of services and building blocks where's the ML action what if I Michelle wanna jump in this community I'm watching this with hey you know what I got Amazon Web Services reinvent just pumping up a lot of MLA I you know stage maker and a bunch of other things what's going on in this community where are the projects what are the notable things where can I jump in and engage what's the what's that what's that map look like I don't know yeah absolutely so obviously I'm pretty biased you know I helped start cube flow we're very very excited about that the cube flows one yeah absolutely but let me speak a little bit more broadly kubernetes gives you this wonderful platform highly scalable incredibly portable and and I can't overstate how valuable that portability is the reality is is that customers have we talked about data a bunch already they have data on Prem they've data in cloud hey cloud B it's everywhere they want to bring it together they want to bring the the training and the inference to where the data is kubernetes solves that for you it gives you portability and lets you abstract away the underlying stuff it gives you great scalability and reliability and it lets you compose these highly complex pipelines together that let you do real training anywhere rather than having to take all your data and move it through cloud and train on a single VM that you're not sure whether or not it's been updated or not this is the way to go versus the old way which was what cuz that's an easier way orchestrating and managing that what was the alternative the alternative was you built it yourself you you piece together a whole bunch of solutions you wired it together you made sure that this service over here had the right user account to access the data that that service over there was outputting it was just a crazy time now you use kubernetes constructs use first-class objects you extend the native kubernetes api and it works on your laptop and it works on Cloud a and B and on pram and wherever you need it that's the magic basically absolutely so multi cloud has come up a lot hybrid clouds the buzzword of the year I call that the 2000 18 maybe 19 buzzword but I think the real end game and all this is what from a customer standpoint that we are reporting a silk'n angle on the cube is choice yeah multi vendor is the new multi cloud is the multi clouds the modern version of the old multi vendor comes yes which basically is choice absolutely so how does kubernetes fit into the multi cloud why is that good for the industry and what's your take on that can you share your perspective absolutely so when you go and look at the recent right scale reports 81 percent of enterprises today are multi cloud . 81 percent and not just one cloud there they're on five different clouds that could be on pram could be multi zone could be Google or Amazon or a Salesforce you name how you define cloud they're spreading they're doing it because that kind of portability is right for their business kubernetes gives you the opportunity to operate in an abstraction layer that works across all of these clouds so whether or not you're on your laptop and you're using docker or mini cube you're on your private training rig whether that you go to Google cloud or as you're on Google clouds you can eat user you have a KS these you're able to build C I'd CD systems continuous delivery systems that that use common kubernetes constructs I want to roll this application out I want there to be seven pods I wanted to have an endpoint that looks like this and that works anywhere you have a kubernetes conformant cluster and when it gets to really complex apps like machine learning you're able to do that it even a higher level using constructs like cube flow and all the many many packages that go into coop load we have Nvidia contributing and we have you know Intel and I mean just countless Cisco I you know I hesitate to keep naming names because I'll be here all day but you know we have literally over Cisco's rays tailwind Francisco they're gonna have Network forever everybody wins at the the CI CD sides for developers one common construct the network guys get more programming because if you decompose an application absolutely the network ties it together yes everybody wins in the stack absolutely I think I breed is really interesting you know hybrid kind of gets a dirty word people like oh my god you know why would you ever deploy to multiple clouds why would you ever spread across multiple clouds and that I agree with a true hybrid deployment today isn't well I'm gonna take my app and I'm gonna spread it across six different locations in fact what you really want to do is have isolated deployments to each place that it enables you in a single button deploy to all three of these locations but to isolate them to have this particular application go and if you know AWS hasn't added GCP is there or if GCB does manage asher is there and you can do that very readily or you can bring it closed for geographic reasons or legal reasons or whatever it might be those kind of flexibility that ability to take a single construct of your application and deploy it to each one of these locations not spreading them but in fact just giving you that flexibility gives you pricing power gives you flexibility and lets you take advantage of the operating model if the if the if the ICD is common and that's the key value right there absolutely right David thanks so much coming on cue as usual great commentary great insight there there from the beginning just final question predictions for 2019 I think kubernetes what's gonna happen in 2019 with kubernetes what's your prediction well III think I think you've heard this message over and over again you're seeing kubernetes become boring and and that is incredibly powerful the the stability the flexibility people are building enormous businesses on top of it but not just that they're also continuing to build things like the the custom resource definition which lets you extend kubernetes in a safe and secure way and that's incredibly important that means you don't have to go and check in code into the main tree in order to make extension you're able to build on top of it and you're seeing more and more businesses build eight solutions customer focus solutions well next time we get together I want to do a drill down on the what the word stack means I heard me say kubernetes stack I'm like yeah I think that you love the stack words let a stack anymore sets the services David thanks so much come on I appreciate it here the queue coverage live here in Seattle for coop con cloud native found I'm John Fourier was too many men we back with more after this short break