 Live from Madrid, Spain. It's theCUBE covering HPE Discover Madrid 2017. Brought to you by Hewlett Packard Enterprise. Welcome back to Madrid, Spain, everybody. This is theCUBE, the leader in live tech coverage. And this is day two of our exclusive coverage of HPE Discover 2017. I'm Dave Vellante with my co-host, Peter Burris. Last night was a great night of customer meetings. We stumbled into the CIO meeting. We were at the- And we're quickly ushered out. We were at the analyst event, and of course we met our good friend, Dr. Tom Bradichitz, at the analyst meeting. This is the man who brought a lot of the IoT initiative into HPE, the general manager of the IoT and systems division. Great to see you again, Dr. Tom. Thanks so much for coming on. Well, thank you, Dave, and Peter. It's great to be here at theCUBE. Great to be here at HPE Discover Madrid. Lots of great things happen in it. I can't wait to tell you about them. So we're very excited to have you on. John Furrier and I interviewed you in the very early days after you came over from your previous company, and you had this sort of vision of bringing the HPE into the intelligent edge. And we're like, okay, this sounds really complicated. You got ecosystem, you got all kinds of technologies that you got to develop, hardware, software, and you're making it happen. It's become a meaningful portion of HPE's business. I know you got a long way to go, but congratulations on the progress so far. Give us the update on where you're at. Well, first of all, thank you for that. I appreciate it. I must give credit to my team. I tell them all the time, if you don't execute and do the work, I'm just a science fiction writer. And the vision has come about and we have real customer deployments. Of course, that's the proof of it. At first, we had no products and no customers. Now we have these products that we'll talk about. We have the customer deployments and we're changing things for businesses at the edge. And again, the edge is just not the data center. And the manufacturing floor, we'll talk about refineries, oil rigs, those type of edges. We're doing a lot of work there. And it's been exciting to see the ideas that we have get adopted by not only customers, but the industry. So we're seeing other analysts pick up on two dimensions, computing at the edge, and a little more complicated one and a little more difficult to grasp is converged OT and IT at the edge, the two worlds of operational technology converging with IT. We were on theCUBE talking with an OT partner, National Instruments, a long while ago. And now we literally have those products in the market, in the hands of customers. National Instruments is reselling the Edge Line 1000, the Edge Line 4000 products, as well as of course us selling it. And it's pretty exciting to see this happening. Well, what I love about that conversation is when we first started to talk to you, we said, okay, let's play the skeptic, analysts are skeptic. And we said, one of the big problems you're going to face is bringing the organizations together, OT and IT. They're just different worlds, oil and water. You know, you got hardcore engineers and you got IT guys. And then subsequent to that conversation, you bring on National Instrument, right? And we have that conversation. Okay, so we said, I checked that box, at least they're having conversations. Can you talk about how that convergence is actually occurring and what's in it for the customer? Well, great. To talk about this convergence, the best thing to do is say it can happen at several levels. It can happen at a solutions level, it can happen at a software level and a hardware physical level. Let's talk about a physical level, it's a little more tangible, understand. Let me use the smartphone, which everybody has. Peter, you have one there. If you hold that up, you will notice inside the manufacturer of that phone converged or integrated those asynonyms, many consumer devices, such as what? A music player, of course, the phone, of course, but also many other things. A GPS system, a camera, the list goes on, right? We can go on the flashlight. And by the way, your wallet, maybe not your wallet, but a millennial and younger's wallet is in that phone. My wallet, too. My wallet, too. Venmo, baby. That's right. My kid's wallet's in there, too. Yeah. Oh, that's great. You've done that switch, so what is happening there, obviously, is the notion of software defining and we're converging. Now, the benefits of that are irrefutable. One thing you buy, it's less energy. One thing to manage, the convenience of carrying around. Let's take that metaphor and impute it at, let me see, a manufacturing floor edge. There's lots of edges out there. We go to a manufacturing floor edge. We see several devices. Just like the early pioneers of the smartphone saw a consumer with a camera around his neck, a GPS on the bell, text, right? A flashlight, a wallet, and all this. We see all these devices out there. And what are they? Some of them are OT, as you mentioned, operational technology devices, such as control systems, such as data acquisition systems. Real-time systems. Real-time systems. Industrial networks. CAN, Profibus, SCADA solutions and networks. The second thing we see is some IT. Most of it's closed, though. This is important. It's good IT, computing and storage, but a lot of it is closed systems. It's not the open X86 architecture that we so enjoy in the data center. So those things are out there. We looked at them and we put them all in one box, just like the smartphone is one device. What are the benefits? Lower space, there's not a lot of space at the edge. Lower energy, there's not a lot of energy right at the edge. But the more profound benefits that we're seeing, and we have a large auto manufacturer who has deployed this on their manufacturing line, is it keeps up time higher. In other words, it reduces downtime. So if the manufacturing line stops, there's nothing worse than a manufacturing line stopped, except perhaps an empty one. But the point is, when a manufacturing line stops, you can't put out product. You can't put out product. You can't recognize revenue you get in the consumer's hands. It's very obvious. It's an airtight business case, actually. So we're able to reduce any downtime. Why? Because first of all, everything's together. And secondly, we're able to manage it just like we're managing the data center, because it's an OpenX86 architect. So you're converging tasks as well as hardware. As well as hardware. And then the next step is software, as well. We just launched a new class of software called the Edge Line Services Platform. And this is OT software. So we're taking OT functions like aggregators and things that do OT technologies and some IT. But because we have so much compute power, and it's open, it's x86, it can run software like VMware, Microsoft products, even database products as well. But because we have that, we're able to software define. When you software define, and I'll use the wallet again, you don't have a billfold with your license anymore, plastic and leather has been software defined, and therefore it's less to deal with. It's much more efficient. So that announcement of our software strategy along now with our hardware strategies is very exciting for us and customers are very much interested in it. So do you have some examples, some real world examples, customers that you can talk about where you're bringing together OT and IT disciplines? Yeah, you bet. Let me talk about a large global beverage and snack company. And they make snacks, and in this case, potato chips. So a potato chip is a product, and the idea of having them come out of the line in the bag and be a higher quality is important. So we took an Edge Line System, the EO1000, and we put it at the edge, and we were able to software define several of their IT and OT components and get into a consolidation, an integration in one box. Now what that did is it allowed the, and will do, is allowed the foods to move faster. So if they move across the conveyor faster, you can bag them faster, get them out to consumer. The second thing is because it's so powerful, this is interesting. Now they can use video cameras to inspect the quality. Now think about that. That's not necessarily a new idea, but what is new is the notion that you can take video, which I think you'd agree is the largest data. Is that right? Video is big, big data. We know that well. Especially if it's high resolution, and your hosting costs are telling you that as well, of all these videos. But if it's high resolution, and because you're looking for defects, indeed one has to process that, not only in high resolution, massive data, number one, number two, quickly, because the thing is moving, and you want to know to knock it off or stop, or whatever the case may be. So what has happened there is, my team and I did not think of that. Our customers thought that well, because you gave us this platform, we can now enhance it with a new type of sensor called a camera, with a new type of data called video to enhance our quality and keep our process moving faster. So keeping this converged notion going, you're converging the hardware, which is important. You're converging a lot of the administrative tasks, which reduces the likelihood of any single human failure bringing the whole system down. But now you're talking about, in the whole sense, infer and act loop that typifies what happens at the edge, you're converging new technologies into that loop by being able to add new data types, bring modeling, machine learning, analytics in the infer, and then being able to act right there, which allows you to think about new invention, new innovation very, very rapidly, because you have the processing power to converge all that new function as it becomes better understood. Have I got that right? You've got it right. I serve as an adjunct professor at a university, so let me position it in an easy way to learn. You said sense infer and act. Let's call them the three A's. Acquire, analyze, and act. It's just easy to remember. And let me talk to it to that, but it's actually a synonyms. So the acquisition of the data is through sensors in D to A conversion, or let me say A to D, analog to digital. Because most of these phenomenon video, for example, has to be, is a light phenomenon. Moisture, pressure, at Duke Energy, for example, the second largest energy provider I worked on that, industrial internet of things solution, and vibration was the thing that needed to be acquired, and then analog to digital. Now the analysis test takes place. There are seven reasons to analyze at the edge. There are seven reasons not to send the data to the cloud. In the past we have talked about it. One of them is latency, one of them is cost, one of them is bandwidth, another one is security, another one is reliability, another one is geofencing and policy, another one is duplication, and security of hostile or just reliability drop packets. There's a lot of issues to do that analysis there. But because we have a non-compromised, full x86, in fact 64 in one box, 64 Zeon, Intel Zeon product, in a one box, we don't have to compromise the stack. We can take it directly out of the data center and run things like artificial intelligence, machine learning algorithms. We can virtualize, we can containerize, we can run Citrix applications at the edge to have better access to the data and of course the application. So but you're absolutely right. And then the second thing in this point is we move from the middle A analysis to the action. The reason, I've learned this, doing many IoT deployments, the reason people do an IoT deployment is to act. Yes, it's exciting to collect data. It's also exciting to analyze it. But have you ever been in a business meeting where you sit and you analyze data and you give tremendous insights and one conclusion is pit against another conclusion and it cancels out all conclusiveness. And then you talk and you analyze and you walk out and nothing happens, there's no action. Many of us have been in that. That's the idea here. You can't stop at the analysis even though artificial intelligence, deep algorithms, moving averages, signatures that we can compare are very powerful. Well, what do you do when you do that? Because we have control and actuation systems built in the edge line, we literally in a physical space as well as a logical process, as you pointed out, close that loop, acquire, analyze, act, acquire, analyze, act, yes, connect to the cloud or the data center if we need to, but the issue is you don't have to. Now here's what's profound about that. This system at the edge can be managed and run the same stacks as in a cloud or data center. I'm going to use those as symptoms because a cloud is just a data center that nobody's supposed to know where it is. So a data center far away on the corporate campus or in a public or private cloud somewhere is managed the same way. When that happens, we are revolutionizing workload management. Now, I spent a lot of years in my former time in IT and building data centers and building some of the first clouds. Workload management's a big deal. How do you shift the workload to the free server or to the free resources, right? To optimize, obviously. It's a packing problem many times in a data center. Well, now we've introduced another place to workload management. It's called the edge. It's far away. So where we workload manage in the data center, then the cloud was invented. That's the first off-premises. The next off-premises is now the edge. So the other off-premises is the edge. So now we have a workload management capability. Do you want to do 100% processing at the edge where the action is and where the acquisition is? Do you want to do 100% in the cloud? That's still possible. Do you want to do 50-50? Would you like to do 10-90? Would you like to do 30-70? You get my point. I can shift this and depending on the season, depending on issues like disaster recovery, depending on your workloads, you can now do that. And again, you can do this with the edge line 1,000, the edge line 4,000 because of the processing power and the converged OT inside it. Well, our observation is that it's not about bringing your business to the cloud. It's about bringing the cloud to your business. So bringing that sense of workload management. You might say the cloud is just a virtualized data center when you come right down to it. So bringing all those capabilities and bringing them to wherever the data requires it. And there's going to be a lot of instances where the data is going to be at the edge, stay at the edge, but that doesn't mean you don't want all the benefits of how you run computing down at the edge where that data is. Yeah, we're not obviating, we're offering choice. But again, there are seven reasons I want to know why you do it here, but I've had a customer say none of those seven matters, okay, we send everything to the cloud and we have great cloud hybrid IT products that do that. And we've envisioned a three tiered data model, real time at the edge, maybe you don't persist everything, but like you said, a lot of reasons not to move all the data back, but there is maybe a spot where you aggregate some of that data from discrete devices. And sure, if you want to do some deep modeling in the cloud, go for it. And that cloud might be the public cloud, it might be your on-prem cloud. Does that seem reasonable to you? Very reasonable. And another reason for a cloud is it's an aggregation point for other, in this case, manufacturing lines or other smart cities to come together because you're not going to connect every city, every plant, you know, any to any, you'll have a hub and spoke cons model where the cloud serves as that hub. So there are always reasons and that's why, you know, if you look at our company, the pillars of our company, point next services, the second pillar is hybrid IT, primarily focused on cloud and data centers, and the third is the intelligent edge. And those all play very, very closely together. In fact, we have edge-to-core strategies, we have edge-to-core offerings with partners like NVIDIA, with partners like SAP, with partners like SAS, we have edge-to-core, for example, Schneider as well, Schneider Electric, all of them are looking at this idea, GE, Microsoft Azure, let's go to the edge. And two years ago, that was not the case, right? Let's go there. When you go to the edge, what are you going to run it on? Well, let's not force our software partners to re-architect like they used to have to to run at the edge, which is like, I'd call that drive-by analytics, you just have to cut out everything because it only ran on a Wimpy Core somewhere or a little device. No, let's move the entire data center capability out to the edge. When I was presenting this to one of our partners, the CEO company, I was presenting this vision and he was texting during my talk because I was boring. And then I said this, this is a very powerful company, we'll miss a name, see it. Then I said, we're going to move data center class technology out to the edge. It's not going to be in compromised cores or limited memory or a little bit of storage. It's the very things in the data center will harden called edge line, will add control systems and data. We'll put it out at the edge. He stopped texting. Maybe looked at me said, wow, you're really moving a data center out to the edge. And you just said that, right? It's the cloud is coming. It's almost the reverse idea of what was happening before. Well, you wrote a blog recently about the space edge. So I wanted to ask you about that. What's going on in the space? And that's the ultimate edge. Yeah, the infinite edge. Explain what you guys are doing there and why it's important. This is exciting. Space travel for exploration and eventually colonization, if you would believe that, is happening. We have the first supercomputer technology in a NASA spaceship now. It has orbited the earth well over a thousand times and it is doing thousands of benchmarks and is doing very well, isn't failing. Now, why is that profound? Because again, that edge is so far away and the ability to push that back to earth now, which we could call the data centers in the earth, is limited. It takes minutes, sometimes even longer. There's issues with reliability as well. So we were able to do that and then we've created a new thing called Project Extreme Edge where we're going to build edge line systems that will fit better with lower energy, smaller size in spaceships, and eventually in colonization. But we're just going into space travel and exploration right now. And I'd like to mention that HPE Labs is a great participant in this because they're working on a technology. And the name of it is called the Dot Product Engine. And Dot Product is a mathematical operation needed in high performance computing in artificial intelligence. But we're able to use that technology because it's small, it's fast, faster than we believe anything else on the market. And also it has a low energy profile. And those are all for any edge, obviously. But it's also great for the space edge. And I like to quote Frank Sinatra when he said, if I can make it there, I can make it anywhere in New York, New York. Well, if we can make it in the space edge, these earth edges will benefit as well. Some of the same challenges. All right, we're out of time, but I got to ask you. Meg stopped by yesterday. And this is given great support for the intelligent edge. Yes, yes. The company's now reporting the intelligent edge is going to be one of the main areas. What about the new guy? Antonio. You know, what's your relationship with him, experience, has he been focused on this area? Support. He's been great. He's support in three ways. Let me just sum up in three ways. Number one, he supports in customer visits. He and I have been on customer visits together. It's always wonderful to have the president and now the new CEO with you affirming what we're doing. That's number one of three. Number two of three, he supports the work we're doing with our new global IoT innovation labs. In fact, our first grand opening, the first one in Houston, we will have one in Singapore opening in February and then we'll have one in Europe and perhaps one in India. We're opening these labs for innovation. But my point is, the one in Houston, our first grand opening, Antonio Neary came personally. He did the ribbon cutting and sponsored that as well. And then third, he is, of course, funding my business unit. And he's been very, very supportive and I'm really happy that he's staying with us and he'll be CEO. Excellent, Dr. Tom, thanks so much for coming on theCUBE. Congratulations, as they say. We know there's a long way to go, but looks like you're off to a great start and have some real traction. So we appreciate your time and your insights. Okay, keep it right there, buddy. We'll be back with our next guest. This is theCUBE. We're live from Madrid. Right back.