>> Narrator: From theCUBE studios in Palo Alto in Boston, connecting with our leaders all around the world. This is theCUBE conversation. >> Welcome back everybody. Jeff Frick here with theCUBE we are in our Palo Alto studios today. We're still getting through COVID, thankfully media was a necessary industry, so we've been able to come in and keep a small COVID crew, but we can still reach out to the community and through the magic of the internet and camera's on laptops, we can reach out and touch base with our friends. So we're excited to have somebody who's talking about and working on kind of the next big edge, the next big cutting thing going on in technology. And that's the internet of things you've heard about it the industrial Internet of Things. There's a lot of different words for it. But the foundation of it is this company it's Intel. We're happy to have joined us Bill Pearson. He is the Vice President of Internet of Things often said IoT for Intel, Bill, great to see you. >> Same Jeff. Nice to be here. >> Yeah, absolutely. So I just was teasing getting ready for this interview, doing a little homework and I saw you talking about Internet of Things in a 2015 interview, actually referencing a 2014 interview. So you've been at this for a while. So before we jump into where we are today, I wonder if you can share, you know, kind of a little bit of a perspective of what's happened over the last five or six years. >> I mean, I think data has really grown at a tremendous pace, which has changed the perception of what IoT is going to do for us. And the other thing that's been really interesting is the rise of AI. And of course we need it to be able to make sense of all that data. So, you know, one thing that's different is today where we're really focused on how do we take that data that is being produced at this rapid rate and really make sense of it so that people can get better business outcomes from that. >> Right, right. But the thing that's so interesting on the things part of the Internet of Things and even though people are things too, is that the scale and the pace of data that's coming off, kind of machine generated activity versus people generated is orders of magnitude higher in terms of the frequency, the variety, and all kind of your classic big data meme. So that's a very different challenge then, you know, kind of the growth of data that we had before and the types of data, 'cause it's really gone kind of exponential across every single vector. >> Absolutely. It has, I mean, we've seen estimates that data is going to increase by about five times as much as it is today, over the next, just a couple years. So it's exponential as you said. >> Right. The other thing that's happened is Cloud. And so, you know, kind of breaking the mold of the old mold roar, all the compute was either in your mini computer or data center or mainframe or on your laptop. Now, you know, with Cloud and instant connectivity, you know, it opens up a lot of different opportunities. So now we're coming to the edge and Internet of Things. So when you look at kind of edge in Internet of Things, kind of now folding into this ecosystem, you know, what are some of the tremendous benefits that we can get by leveraging those things that we couldn't with kind of the old infrastructure and our old way kind of gathering and storing and acting on data? >> Yeah. So one of the things we're doing today with the edge is really bringing the compute much closer to where all the data is being generated. So these sensors and devices are generating tons and tons of data and for a variety of reasons, we can't send it somewhere else to get processed. You know, there may be latency requirements for that control loop that you're running in your factory or there's bandwidth constraints that you have, or there's just security or privacy reasons to keep it onsite. And so you've got to process a lot of this data onsite and maybe some estimates or maybe half of the data is going to remain onsite here. And when you look at that, you know, that's where you need compute. And so the edge is all about taking compute, bringing it to where the data is, and then being able to use the intelligence, the AI and analytics to make sense of that data and take actions in real time. >> Right, right. But it's a complicated situation, right? 'Cause depending on where that edge is, what the device is, does it have power? Does it not have power? Does it have good connectivity? Does it not have good connectivity? Does it have even the ability to run those types of algorithms or does it have to send it to some interim step, even if it doesn't have, you know, kind of the ability to send it all the way back to the Cloud or all the way back to the data center for latency. So as you kind of slice and dice all these pieces of the chain, where do you see the great opportunity for Intel, where's a good kind of sweet spot where you can start to bring in some compute horsepower and you can start to bring in some algorithmic processing and actually do things between just the itty-bitty sensor at the itty-bitty end of the chain versus the data center that's way, way upstream and far, far away. >> Yeah. Our business is really high performance compute and it's this idea of taking all of these workloads and bringing them in to this high performance compute to be able to run multiple software defined workloads on single boxes, to be able to then process and analyze and store all that data that's being created at the edge, do it in a high performance way. And whether that's a retail smart shelf, for example, that we can do realtime inventory on that shelf, as things are coming and going, or whether it's a factory and somebody's doing, you know,real time defect detection of something moving across their textile line. So all of that comes down to being able to have the compute horsepower, to make sense of the data and do something with it. >> Right, right. So you wouldn't necessarily like in your shelf example that the compute might be done there at the local store or some aggregation point beyond just that actual, you know, kind of sensor that's underneath that one box of tide, if you will. >> Absolutely. Yeah, you could have that on-prem, a big box that does multiple shelves, for example. >> Okay, great. So there's a great example and you guys have the software development kit, you have a lot of resources for developers and in one of the case studies that I just wanted to highlight before we jump into the dev side was I think Audi was the customer. And it really illustrates a point that we talked about a lot in kind of the big data meme, which is, you know, people used to take action on a sample of data after the fact. And I think this case here we're talking about running 1,000 cars a day through this factory, they're doing so many welds, 5 million welds a day, and they would pull one at the end of the day, sample a couple welds and did we have a good day or not? Versus what they're doing now with your technology is actually testing each and every weld as it's being welded, based on data that's coming off the welding machine and they're inspecting every single weld. So I just love you've been at this for a long time. When you talk to customers about what is possible from a business point of view, when you go from after the fact with a sample of data, to in real time with all the data, how that completely changes your view and ability to react to your business. >> Yeah. I mean, it makes people be able to make better decisions in real time. You know, as you've got cameras on things like textile manufacturers or footwear manufacturers, or even these realtime inventory examples you mentioned, people are going to be able to make and can make decisions in real time about how to stock that shelf, what to order about what to pull off the line, am I getting a good product or not? And this has really changed, as you said, we don't have to go back and sample anymore. You can tell right now as that part is passing through your manufacturing line, or as that item is sitting on your shelf, what's happening to it. It's really incredible. >> So let's talk about developers. So you've got a lot of resources available for developers and everyone knows Intel obviously historically in PCs and data centers. And you would do what they call design wins back when I was there, many moons ago, right? You try to get a design win and then, you know, they're going to put your microprocessors and a bunch of other components in a device. When you're trying to work with, kind of Cutting Edge Developers in kind of new fields and new areas, this feels like a much more direct touch to the actual people building the applications than the people that are really just designing the systems of which Intel becomes a core part of. I wonder if you could talk about, you know, the role developers and really Intel's outreach to developers and how you're trying to help them, you know, kind of move forward in this new crazy world. >> Yeah, developers are essential to our business. They're essential to IoT. Developers, as you said, create the applications that are going to really make the business possible. And so we know the value of developers and want to make sure that they have the tools and resources that they need to use our products most effectively. We've done some things around OpenVINO toolkit as an example, to really try and simplify, democratize AI application so that more developers can take advantage of this and, you know, take the ambitions that they have to do something really interesting for their business, and then go put it into action. And the whole, you know, our whole purpose is making sure we can actually accomplish that. >> Right. So let's talk about OPenVINO. It's an interesting topic. So I actually found out what OpeVINO means, Open Visual Inference and Neural Optimization toolkit,. So it's a lot about computer vision. So I will, you know, and computer vision is an interesting early AI application that I think a lot of people are familiar with through Google photos or other things where, you know, suddenly they're putting together little or a highlight movies for you, or they're pulling together all the photos of a particular person or a particular place. So the computer vision is pretty interesting. Inference is a special subset of AI. So I wonder, you know, you guys are way behind OpenVINO. Where do you see the opportunities in visualization? What are some of the instances that you're seeing with the developers out there doing innovative things around computer vision? >> Yeah, there's a whole variety of used cases with computer vision. You know, one that we talked about earlier here was looking at defect detection. There's a company that we work with that has a 360 degree view. They use cameras all around their manufacturing line. And from there, they didn't know what a good part looks like and using inference and OpenVINO, they can tell when a bad part goes through or there's a defect in their line and they can go and pull that and make corrections as needed. We've also seen, you know, use cases like smart shopping, where there's a point of sale fraud detection. We call it, you know, is the item being scanned the same as the item that is actually going through the line. And so we can be much smarter about understanding retail. One example that I saw was a customer who was trying to detect if it was a vodka or potatoes that was being scanned in an automated checkout system. And again, using cameras and OpenVINO, they can tell the difference. >> We haven't talked about a computer testing yet. We're still sticking with computer vision and the natural language processing. I know one of the areas you're interested in and it's going to only increase in importance is education. Especially with what's going on, I keep waiting for someone to start rolling out some national, you know, best practice education courses for kindergartens and third graders and sixth graders. And you know, all these poor teachers that are learning to teach on the fly from home, you guys are doing a lot of work in education. I wonder if you can share, I think your work doing some work with Udacity. What are you doing? Where do you see the opportunity to apply some of this AI and IoT in education? >> Yeah, we launched the Nanodegree with Udacity, and it's all about OpenVINO and Edge AI and the idea is, again, get more developers educated on this technology, take a leader like your Udacity, partner with them to make the coursework available and get more developers understanding using and building things using Edge AI. And so we partnered with them as part of their million developer goal. We're trying to get as many developers as possible through that. >> Okay. And I would be remiss if we talked about IoT and I didn't throw 5G into the conversation. So 5G is a really big deal. I know Intel has put a ton of resources behind it and have been talking about it for a long, long time. You know, I think the huge value in 5G is a lot around IoT as opposed to my handset going faster, which is funny that they're actually releasing 5G handsets out there. But when you look at 5G combined with the other capabilities in IoT, again, how do you see 5G being this kind of step function in ability to do real time analysis and make real time business decisions? >> Well, I think it brings more connectivity certainly and bandwidth and reduces latency. But the cool thing about it is when you look at the applications of it, you know, we talked about factories. A lot of those factors may want to have a private 5G networks that are running inside that factory, running all the machines or robots or things in there. And so, you know, it brings capabilities that actually make a difference in the world of IoT and the things that developers are trying to build. >> That's great. So before I let you go, you've been at this for a while. You've been at Intel for a while. You've seen a lot of big sweeping changes, kind of come through the industry, you know, as you sit back with a little bit of perspective, and it's funny, even IoT, like you said, you've been talking about it for five years and 5G we've been been waiting for it, but the waves keep coming, right? That's kind of the fun of being in this business. As you sit there where you are today, you know, kind of looking forward the next couple of years, couple of four or five years, you know, what has just surprised you beyond compare and what are you still kind of surprised that's it's still a little bit lagging that you would have expected to see a little bit more progress at this point. >> You know, to me the incredible thing about the computing industry is just the insatiable demand that the world has for compute. It seems like we always come up with, our customers always come up with more and more uses for this compute power. You know, as we've talked about data and the exponential growth of data and now we need to process and analyze and store that data. It's impressive to see developers just constantly thinking about new ways to apply their craft and, you know, new ways to use all that available computing power. And, you know, I'm delighted 'cause I've been at this for a while, as you said, and I just see this continuing to go far as far as the eye can see. >> Yeah, yeah. I think you're right. There's no shortage of opportunity. I mean, the data explosion is kind of funny. The data has always been there, we just weren't keeping track of it before. And the other thing that as I look at Jira, Internet of Things, kind of toolkit, you guys have such a broad portfolio now where a lot of times people think of Intel pretty much as a CPU company, but as you mentioned, you got to FPGAs and VPUs and Vision Solutions, stretch applications Intel has really done a good job in terms of broadening the portfolio to go after, you know, kind of this disparate or kind of sharding, if you will, of all these different types of computer applications have very different demands in terms of power and bandwidth and crunching utilization to technical (indistinct). >> Yeah. Absolutely the various computer architectures really just to help our customers with the needs, whether it's high power or low performance, a mixture of both, being able to use all of those heterogeneous architectures with a tool like OpenVINO, so you can program once, right once and then run your application across any of those architectures, help simplify the life of our developers, but also gives them the compute performance, the way that they need it. >> Alright Bill, well keep at it. Thank you for all your hard work. And hopefully it won't be five years before we're checking in to see how far this IoT thing is going. >> Hopefully not, thanks Jeff. >> Alright Bill. Thanks a lot. He's bill, I'm Jeff. You're watching theCUBE. Thanks for watching, we'll see you next time. (upbeat music)