 Live from Las Vegas, it's theCUBE. Covering Discover 2016 Las Vegas. Brought to you by Hewlett Packard Enterprise. Now, here are your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live in Las Vegas for SiliconANGLE Media's theCUBE. It's our flagship program. We go out to the events to extract the signal of noise. We hear exclusive coverage of HP Enterprise Discover 2016. I'm John Furrier with my co-host Dave Vellante extracting the signal of noise with two great guests, Dr. Tom Bradis, VP and general manager of the servers and IoT systems, and Eric Starkoff, the stockloft, the EVP of global sales and marketing at National Instruments. Welcome back to theCUBE. Thank you. Welcome on for the first time. Cube alumni, welcome to theCUBE. Thank you. So, we are seeing a real interesting historic announcement from HP because not only is there an IoT announcement this morning that you are the architect of, but the twist that you're taking with IoT is very cutting edge. Kind of like I was just at Google I.O. and these big comments. They always have some sort of sexy demo that to kind of show the customers the future, like AI or, you know, Oculus Rift goggles as the future of their application. You actually don't have something that's futuristic. It's reality. You have a new product around IoT at the edge. Edge line, the announcements are all online. Tom, but you guys did something different. And Eric's here for a reason. We'll get to that in a second, the announcement represents a significant bet that you're making and HP's making on the future of IoT. Please share the vision and the importance of this event. Well, thank you. And it's great to be here and back here with you guys. We've looked around and we could not find anything that existed today, if you will, to satisfy the needs of this industry and our customers. So we had to create not only a new product, but a new product category. A category of products that didn't exist before and the new edge line 1000 and the edge line 4000 are the first entrance into this new product category. Now what's a new product category? Well, whoever invented the first automobile, there was not a category of automobiles. When the first automobile was invented, it created a new product category called automobiles and today everybody has a new entry into that as well. So we're creating a new product category called converged IoT systems. Converged IoT systems are needed to deliver the real time insights, real time response and advance the business outcomes or the engineering outcomes or the scientific outcomes, depending on the situation of our customers. They're needed to do that. Now, when you have a name converged, that means in somewhat a synonym is a integration. What did we integrate? And I want to tell you the three major things we integrated, one of which comes from Eric and the fine national instruments company that makes this technology that we actually put in to the single box. And I can't wait to tell you more about it, that's what we did, a new product category, not just two new products. And so you guys are bringing two industries together. Again, that's not only just point technologies or platforms, tooling, you're bringing disparate kind of players together. But it's not just a partnership, it's not like shaking hands and doing a strategic partnership. So there's real meat on the bone here. Eric, talk about one, the importance of this integration of two industries basically coming together, converged category, if you will, or industry and what specifically is in the box or in the technology? Yeah, I think you hit it exactly right. I mean, everyone talks about the convergence of OT or operational technology and IT, and we're actually doing it together. I represent the OT side, national instruments as a global leader. OT is means just for the audience. Operational technology, it's basically industrial equipment, measurement equipment, the thing that is connected to the real world, taking data and controlling the thing that is in the internet of things or the industrial internet of things as we play. And IT is infrastructure technologies, we know that is. OT is operational technology. We've been doing IoT before it was a buzzword, doing measurement and control systems on industrial equipment. So when we say we're making it real, this edge line system actually incorporates in national instruments technology on an industry standard called PXI. And it is a measurement and control standard that's ubiquitous in the industry and it's used to connect to the real world, to connect to sensors, actuators, to take in image data and temperature data and all of those things to instrument the world and take in huge amounts of analog data and then apply the compute power of an edge line system onto that application. We don't talk a lot about analog data in the IT world. Why is analog data so important? I mean, it's prevalent, obviously, in your world. Talk a little bit more about that. It's the largest source of data in the world, as Tom says, it's the oldest as well. Analog, of course, if you think about it, the analog world is literally infinite and it's only limited by how many things we want to measure and how fast we measure them. And the trend in technology is more measurement points and faster. Let me give you a couple of examples of the world we live in. Our customers have acquired over the years approximately 22 exabytes of data. We don't deal with exabytes that often, I'll give an analogy. It's streaming high definition video continuously for a million years, produces 22 exabytes of data. Customers like CERN that do the Large Hadron Collider, they're a customer of ours, they take huge amounts of analog data. Every time they do an experiment, it's the equivalent of 14 million images, photographs that they take per second. They create 25 petabytes of data each year. The importance of this and the importance of Edgeline, and we'll get into this some, is that when you have that quantity of data, you need to push processing and compute technology towards the Edge for two main reasons. One is the quantity of data doesn't lend itself or takes up too much bandwidth to be streaming all of it back to central, to cloud or centralized storage locations. The other one that's very, very important is latency. In the applications that we serve, you often need to make a decision in microseconds. And that means that the processing needs to be done, literally the speed of light is a limiting factor. The processing must be done on the Edge at the thing itself. So basically you need a data center at the Edge. Exactly. A great way to say it. A great way to say it. And this data or big analog data, as we love to call it, is things like particulates, motion, acceleration, voltage, light, sound, location, such as GPS, as well as many other things like vibration and moisture. That is the data that is pent up in things, in the Internet of Things. And Eric's company, National Instruments, can extract that data, digitize it, make it ones and zeros, and put it into the IT world where we can compute it and gain these insights and actions. So we really have a seminal moment here. We really have the OT industry represented by Eric, connecting with the IT industry in the same box, literally in the same product in the box. Not just a partnership, as you point out. In fact, it's quite a moment. I think we should have a photo op here, shaking hands. Two industries coming together. So you talked about this new product category. What are the parameters of a new product category? You gave an example of an automobile. Okay, but nobody had ever seen one before. But now you're bringing together two worlds. What defines the parameters of a product category such that it warrants a new category? In general, never been done before and accomplishes something that's not been done before. So that would be more general. But very specifically, this new product, Edge Line 1000 and 4000, creates a new product category because this is an industry first. Never before have we taken data acquisition and captured technology from national instruments and data control technology from national instruments, put that in the same box as deep compute, deep x86 compute. What do I mean by deep? 64 Xeon cores, as you said, a piece of the data center. But that's not all we converged. We took Enterprise Class Systems Management, something that HP has done very well for many, many years. We've taken the Hewlett-Packard Enterprise ILO, Lights Out technology, converged that as well. In addition, we put storage in there. Tens of terabytes of storage can be at the edge. So by this combination of things, that did exist before, the elements of course, by that combination of things, we've created this new product category. And is there a data store out there as well or database? Oh yes, now, since we have, this is the profundity of what I said lies in the fact that because we have so many cores so close to the acquisition of the data from national instruments, we can run virtually any application that runs on an x86 server. So, and I'm not exaggerating, thousands, thousands of databases, machine learning, manageability, insight, visualization of data, data capture tools that all run on servers and workstations now run at the edge. Again, that's never been done before in the sense that at the edge today are very weak processing, very weak. And you can't just run an unmodified app at that level. And in terms of the value chain national instrument is a supplier to this new product category? Is that the right way to think about it or is it? An ingredient, a solution ingredient, but just like we are, number one, but we are both reselling the product together. So we've jointly, collaboratively developed this together. So it's engineers and engineers getting together, building the product? It's engineers mind, we worked extremely close and produced this beauty. We had a conversation yesterday, argument about the iPhone. I was saying, hey, this was a game changing category, if you will, because it was a computer that had software that could make phone calls. Versus the other guys who had a phone that could do text messages and do email. That's a great way to look at that converge product. This would be similar if I may, if you can correct me if I'm wrong, I want you to correct me and clarify. What you're saying is you guys essentially looked at the edge differently saying, let's build the data center at the edge in theory or in concept here, in a little concept, but in theory, the power of a data center that happens to do edge stuff. That's right. Is that accurate? I think it's very accurate. Let me make a point and let you respond. Neapolitan ice cream has three flavors, chocolate, vanilla, strawberry, all in one box. That's what we did with this edge line. What's the value of that? Well, you can carry it, you can store it, you can serve it more conveniently with everything together. You could have separate boxes of chocolate, vanilla, strawberry, that existed, right? But coming together, that convergence is key. We did that with deep compute, with data capture and control, and then systems management, enterprise class device and systems management. And I'd like to explain why this is a product. Why would you use this product as well? And before I continue though, I want to get to the seven reasons why you would use this. We'll go fast, but seven reasons why. But would you like to add anything about the definition of convergence? Yeah, I was going to just give a little perspective on, from an OT, an industrial OT kind of perspective, this world has generally lived in a silo away from IT. It's been proprietary networking standards, not being connected to the rest of the enterprise. That's the huge opportunity, when we talk about the IoT or the industrial IT, is connecting that to the rest of the enterprise. Let me give you an example. One of our customers is Duke Energy. They've implemented an online monitoring system for all of their power generation plants. They have 2,000 of our devices called Compact Rio that connect to 30,000 sensors across all of their generation plants, getting real time monitoring, predictive analytics, predictive failure, and it needs to have processing close to the edge. That latency issue I mentioned, they need to basically be able to do deep processing and potentially shut down a machine. Immediately if it's in a condition that warrants so. The importance here is that as those things are brought online into IT infrastructure, the importance of deep compute, and the importance of the security and the capability that HPE has becomes critical to our customers and the industrial internet of things. Well, I want to push back on, just please devil's advocate and kind of poke holes in your thesis, if I can. So, you got the probes and all the sensors and all the analog stuff that's been going on for years and years, powering instrumentation, you got the box, so okay, I'm a customer. I have other stuff I might put in there, so I don't want to just rely on just your two stuff, your technologies. So how do you deal with the corner case of I might have my own different devices, it's connected through IT, is that just a requirement on your end or is that, how do you deal with that multi-vendor thing? It has to be an open standard and there's two elements of open standard in this product, I'll let Tom comment on one, but one of them is the actual IO standard that connects to the physical world, we said it's something called PXI. National Instruments is a major vendor within this PXI market, but it is an open standard, there are 70 different vendors, thousands of products, so that part of it in connecting to the physical world is built on an open standard and the rest of the platform is as well. Indeed, can I go back to your metaphor of the smartphone that you held up? There are times even today, but it's getting less and less, that people still carry on a camera or a second phone or a music player or the beats, you know, headphones, et cetera, right? There's still time for that, so to answer your question, it's not a replacement for everything, but very frankly, the vision is over time just like the smartphone and the app store, more and more we'll get converged into this platform. So it's a introduction of a platform, we've done the inaugural convergence of the aforementioned data capture, compute, management, storage, and we'll continue to add more and more, again, just like the smartphone analogy and there'll still be peripheral solutions around to address your point. So your multi-vendor strategy, if I get this right, doesn't prevent you, doesn't foreclose the customer's benefits in any way, so they connect through IT, they're connected into the box and benefits, you today did just not converge inside the box. At this point, but I'm getting calls regularly and you may too, Eric, of other vendors saying I want in and I would like to relate that conceptually to the app store, right? Third-party apps are being produced all the time that go on to this platform and it's pretty exciting. Before you get to your seven killer attributes, what's the business model? So you guys have jointly engineered this product, jointly selling it through your channels, if you have large customer like GE, for example, who just sort of made the public commitment to HPE infrastructure, how will you guys split the booty, so to speak? Well, we are actually, as Tom said, we are doing reselling, we'll be reselling this through our channel, but I think one of the key things is bringing together our mutual expertise, because when we talk about convergence of OT and IT, it's also bringing together the engineering expertise of our two companies. We really understand acquiring data from the real world, controlling industrial systems. HPE is the world leader in IT technology, and so we'll be working together and mutually at customers to bring those two perspectives together and we see huge opportunity in that. Yeah, okay, so it's engineering, you guys are channeled, primarily channeled company anyway, so. I can make it frankly real simple, knowing that if we go back to the Neapolitan ice cream and we reference National Instruments Chocolate, they have all the contact with the chocolate vendor, chocolate customers, if you will. We have all the vanilla, so we can go in and then pull each other that way and then go and pull it this way, right? So that's one way, as this market develops. That's going to be very powerful, because indeed, the more we talk about when it used to be separated before today, the more we're expressing that also separate customers that the other guy does not know, and that's the key here in this relationship. So talk about the trend that we're hearing here at the show. I mean, it's been around in IT for a long time, but more now with the agility, the DevOps and cloud and everything, end-to-end management, because that seems to be the stable stakes. Do you address that in the announcement? Is it part of the fit right in? Absolutely, because when we take and we shift left, this is one of our monikers, we shift left, the data center and the cloud's on the right, and we're shifting left, the data center class capabilities out to the edge. That's why we call it shift left. And we meet our partner in national instruments who's already there and an expert in the leader. As we shift left, we're also shifting with it the manageability capabilities and the software that runs management, whether it be infrastructure. I mean, I can do virtualization at the edge now with very popular virtualization packages. I can do remote desktops like the Citrix company, the VMware company, these technologies and databases that come from our own Vertica database that come from PTC, a great partner, again, operations technology. Things that were running already in the data center now get to run there. So you bring the benefit to the IT guy out to the edge to manage and to end. Eric, you get the benefit of connecting into IT to bring that data benefits into the business processes. Exactly, and as the industrial internet of things scales to billions of machines that have monitoring and online monitoring capability, that's critical. It has to be manageable. You have to be able to have these IT capabilities in order to manage such a diverse set of assets. Well, the big data group basically validates that and the whole big data thesis is moving data where it needs to be and having data about physical analog stuff, assets, can come in and surface more insight. Exactly, the big initiative of all. We've got to get to the significant seven. We only have a few minutes left. Hit it. Yeah, we're cliffhanging here on that one. But let me go through them real quick. So the question is, why wouldn't I just, rudimentary collect the data, do some rudimentary analytics, send it all to the cloud? In fact, you hear that today a lot, popular, censored cloud, censored cloud, who doesn't have a cloud today? Every time you turn around, somebody's got a cloud, please send me all your data. We do that and we do that well. We have Helion, we have the Microsoft Azure IoT cloud, we do that well. But my point is there's a world out there and it can be as high as 40 to 50% of the market. IDC is quoted as suggesting 40% of the data collected at the edge, by for example, national instruments, will be processed at the edge, not sent necessarily back to the data center cloud. Okay, with that background, there are seven reasons to not send all the data back to the cloud. It doesn't mean you can or you shouldn't, it just, you don't have to. There are seven reasons to compute at the edge with an edge line system. Ready? Ready. We're gonna go fast and there'll be a test on this. I'm writing it down. Number one is latency. Eric already talked about that. How fast you wanna turn around time? How fast would you like to know your asset's gonna catch on fire? How fast would you like to know in the future, autonomous car that there's a little girl playing in the road as opposed to a plastic bag being blown against the road and are you gonna rely on the latency of going all the way to the cloud and back, which by the way may be dropped as far, not only slow but you're trying to make a phone call recently and it not work, right? So you get that point. So that's latency one. You need time to insight, time to response. Number one of seven, I'll go real quick. Number two of seven is bandwidth. If you're gonna send all this big analog data, the oldest, the fastest and the biggest of all big data all back, you need tremendous bandwidth. And sometimes it doesn't exist or as in some of our mutual customers tell us, it exists but I don't wanna use it all for edge data coming back. That's two of seven. Three of seven is cost. If you're gonna use the bandwidth, you gotta pay for it. Even if you have money to pay for it, you might not want to use it. So again, that's three, let's go to four. Excuse me. Number four of seven is threats. If you're gonna send all the data across sites, you have threats. Doesn't mean we can't handle the threats. In fact, we have the best security in the industry with our Rubik's security, ClearPass. We have ArcSight, we have Volt, we have several things. But the point is, again, it just exposes it to more threats. I've had customers say we don't wanna expose it. Anyway, that's four. Let's move on to five is duplication. If you're gonna collect all the data and then send it all back, you're gonna duplicate at the edge, you're gonna duplicate not all things, but some things both. All right, so duplication. And here we're coming up to number six. Number six is corruption. Not hostile corruption, but just packets drop. Data gets corrupt. The longer you have it in motion, e.g. back to the cloud, right? The longer it is as well. So you have a corruption you can avoid. And number three, I'm sorry, number seven. Here we go with number seven. Not to send all the data back is what we call policies and compliance, geofencing. I've had a customer say I am not allowed to send all the data to these data centers or to my data scientists because I can't leave country borders. I can't go over the ocean as well. Now again, all these seven create a market for us so we can solve these seven or at least significantly ameliorate the issues by computing at the edge with the edge line systems. Great, Eric, I want to get your final thoughts here and let's be wind down the segment. You're from the opposite side, opposite technologies, this is your world. It's not new to you, this edge stuff. It's been there. It's been there done that. It's not, it is IoT for you, right? So you've seen the evolution of your industry. For the folks that are in IT that HP is going to be approaching with this new category and this new shift left, what does it mean? Share your color behind and reasoning and reality check on the viability and relevance. Yeah, I think that there's some significant things that are driving this change. The rise of software capability, the connecting these previously siloed, unconnected assets to the rest of the world is a fundamental shift and the cost point of acquisition technology has come down to the point where we literally have a better, more compelling economic case to be made for the online monitoring of more and more machine type data. That example I gave of Duke Energy, 10 years ago they evaluated online monitoring and it wasn't economical to implement that type of a system. Today it is and it's actually very, very compelling to their business in terms of schedule downtime, maintenance cost, it's a compelling value proposition. And the final one is as we deliver more analytics capability to the edge, I believe that's going to create opportunity that we don't even really completely envision yet and this deep computing that the edge line systems have is going to enable us to do an analysis at the edge that we've previously never done and I think that's going to create whole new opportunities. So based on your expert opinion, talking to the IT guys watching, viability and the ability to do this, what's the, because some people are a little nervous well the parish should open. I mean it's a huge, huge endeavor for an IT company to instrument the edge of their business, it's the bleeding edge, literally. What's the viability, the outcome, is it possible? You see it. It is here now. I mean this announcement kind of codifies it in a product, a new product category, but it's here now and it's inevitable. Final word, your thoughts. I agree. Crowd papa, you're like crowd papa now, you're like your baby out there. It's great, but the more I tell you how wonderful the edge line 1,000, 4,000 is, it's like my mother calling me handsome. Therefore I want to point the audience to FlowServe, F-L-O-W-S-E-R-V-E, they're one of our customers using edge line and national instruments equipment, so we can find that video online as well, but they'll tell us about really the value here and availability is. Right now we have edge line 1,000s and 4,000s in the hands of our customers doing evaluations at the end of the summer. Free announcement, not general availability. Right, general availability is not yet, but we'll have that at the end of the summer and we can do limited availability as we call it depending on the demand and how we roll it out, so. How big the customer base is and relevance to the, is this the old moonshot box, just a quick final question? It is not, no. We are leveraging some high performance low power technology that Intel has just announced. I'd like to shout out to that partner. They just announced and launched Diane Bryant, did her keynote to launch the new Xeon E3, low power high performance Xeon and it was streamed, their keynote on the edge line compute engine. That's actually going into the edge line. That compute blade is going into the edge line. She streamed with it. We're pretty excited about that as well. Tom and Eric, thanks so much for sharing the big news and of course congratulations, new category. Let's see how this plays out. We'll be watching. Gotta get the draft picks in to fit this new sports league. We're calling it like the IoT of the edge. Of course, we're the cube. We're living at the edge all the time. We're at the edge of HP Discover. One more day tomorrow, but again, three days of coverage. You watching theCUBE. I'm John Furrier with Dave Vellante. We'll be right back.