 Live from the San Jose McKennery Convention Center, it's theCUBE at Open Compute Project US Summit 2015. Okay, welcome back everyone. We are live here in Silicon Valley for the Open Compute Project Summit 2015. Hashtag OCP 2015, OCP Summit 2015. This is theCUBE, our flagship program. We go out to the events and extract the civil noise. I'm John Furrier with my co-host, Jeff Frick with theCUBE. Our next guest is Regene Skillern, who's with the general manager of the cloud service, provide a business unit at Intel. Welcome back to theCUBE, great to see you. Nice to see you again. So Intel always been staple in all these super alpha tech athlete events where there's a lot of hardware involved. I mean, Intel inside, you guys are obviously, everyone knows Intel. But now the innovation scale, and you guys have been talking about this for some time. Moore's Law is famous for its, you know, smaller, faster, cheaper, doubling all this stuff. But now you have a whole land grab going on with cloud, density, smaller form factors in the data center, and software on the actual hardware. In memory analytics is a big trend we're seeing across the board, some of the big, you know, in the big data space, Internet of Things. But now we're in the data center. What's this going on? Why is the software component really important? And how does that take advantage to the new hardware coming out? You know, so what we've learned is, you know, decades of creating technology. And our Xeon product line was really optimized for the volume market. The best performance, the best efficiency per dollar in a standard package. But with the cloud, with a lot of these workloads growing rapidly, what we've found is to really push the scale and density and optimization that folks like the cloud need to see in hyperscale, we need to even more finely tune the architecture for the unique workloads that they're running. So that could be things like where we've taken Xeon and actually kind of bifurcated the product line to create our atom roadmap or our FI road map, specifically for certain workloads. But what we've also done is within all of our processor lines as well is integrate a new technology so that the hardware actually works better with the software. This can be accelerators, new instructions and things that we put in to optimize around. But when you talk about in an environment, a truly software divine environment where the application is now in control to pull compute network or storage resources, we want to make sure that they're not just pulling the best compute, but optimized compute, optimized network and storage silicon that's been specifically designed for those unique workload requirements. So we love the focus on innovation and hardware, we love what we're seeing here today. We also love how it marries with the open movement because we do believe that the best form of innovation is when it's built on open standards and that really accelerates the market. So talk about some of the trends here, obviously HP's getting up on stage, they're a big partner of Intel in the past and now they're acknowledging the open source concept of designing in the open. Certainly with open source software, you're starting to see more of this go on and there's this operating system thing going on with the data center orchestrating software. It's kind of a DevOps meets cloud play but now you're talking about like really geeky hardware stuff. So for the folks out there watching CIO or someone in the enterprise who want to be more Facebook-like, how do they get there? What are you guys doing? What is the innovation? What's the enabler right now in this innovation area? Well it does, I'm going to go back to the workload. I think the real innovation and if you look at any of the global cloud service providers out there, one thing they have an advantage on is usually they get to optimize around a fewer set of applications, right? It is their business. And that's why they've custom designed this hardware. When you look at the players that are emerging now like HP's cloud line strategy or some of the many other vendors here with products, what they're trying to do is say, okay, I know that there's other customers who want those same advantages but they maybe don't have the ability to build out of three data centers designed for one application. So what they've done is they've taken the best of scale and flexibility, married in open APIs or open management interfaces. Down from some of these are bare metal servers to the higher levels of service support because once again every customer has a unique set of skills and capabilities to implement and deploy. So by kind of taking that range and really that old traditional, not old but the traditional enterprise class server that's really baked and loaded and robust for an enterprise environment, be able to strip it back and give people that flexibility is kind of marrying that innovation that you were talking about to allow optimization for a unique environment. So do you think that's going to translate into the enterprise? Because that's always a big knock, right? Enterprise people say, well, of course, Facebook doesn't have that many applications, it was a massive pass of scale for not that many applications. And they've got the customized stuff but how does that translate back into the enterprise? But I think it's really interesting that we're starting to see that here. For example, the FSI panel, typically kind of an enterprise class deployment. Not necessarily, I mean, many of their environments are massively scaled but it's that acknowledgement of the by workload and I do think that as enterprise becomes, as they get on this path to software defined infrastructure where they are really focused on TCO optimization or better VM density or higher utilization rates across their infrastructure, they are going to start to really, you know, question the true optimization of their deployments. They're still going to need the management robust frameworks, they're going to need a lot of these things because it provides a lot of value but they're going to have more choices which is interesting. So talk about the workflows. This is an interesting point because now we're talking about the enterprise. Let's go back to the service provider because Mobile World Congress was last week and you hear all about over the top Netflix, all this good stuff happening. We've been watching, you guys have been big supporters of as well. This new era of computing. So I'm a service provider. How does this work for me? I want to be faster. I want to have virtualization. I want to move stuff to the edge of the network. I want low latency. It's all converging in, right? Converged infrastructure. Where's this all mean for those guys? What's their upside here? What's the benefits to say a big service provider? Well, I think for the service providers they have two things they have to focus on. One, they have to get maximum efficiency out of every bit of infrastructure they deploy. It's their cost of sales. So maintaining the dollar total cost of ownership and keeping that cost effective especially, you know, there's hundreds of service providers out there and a lot of them are smaller and they have to compete with the bigger players. They might not have the volume economics. So they have to get really focused on their infrastructure to make sure that they aren't wasteful, that they are highly utilized, that it is specifically built for their environment. But also, you can't just compete on cost alone. It would, you know, that's a commodity environment, not an interesting cloud environment. So what we see is a lot of service providers designing for differentiation and service. And this is a lot of focus that I've seen among our customers where they say, I'm going to get the best out of what I have. Now I'm going to take it the next step and I'm going to optimize around security. How can I enable better security solutions? How can I enable whether that's faster encryption or geo-tagging and secure trust as, you know, to manage some of the government policies or regulation that's out there. Also, people are differentiating around high performance or by workload, HPC clusters in the cloud, big data analytics in the cloud. So I think, you know, once again, it comes back to workload, but it's how these service providers compete. And it also- Mobile's putting pressure on them. Mobile's putting pressure, but it's also creating a huge opportunity for, you know, for every, what is it? Every couple hundred smartphones, you need another data center, or a server in the data center, right? For every 40 smart signs, you need another server in the data center. This whole spiral of more devices, more people using it, more complex data. So do you see the cloud market and the service provider level taking a long tail distribution, meaning you're going to have the big whales and then as you move down into a set of services, you're going to have clouds for everything from dry cleaning to refrigerator maintenance to, you know, whatever, right? I mean, you could have, you know, any level of cloud-based service. We're seeing that SaaS kind of model. You're right. There's a, we talk about, we internally didn't tell, we talked about Jevons Paradox or Jevons Law if you've heard of this. It's about when you can make the cost of something easier to deploy and cheaper, it doesn't create a commodity, it doesn't stop people from buying, they don't buy less. What they do is they actually find new uses for it because it becomes more easily to deploy and more cost-effective. We believe that case with technology. So, you know, as people, as the cloud becomes more cost-effective, as people deploy software-defined infrastructure and really get their cost for operating the infrastructure, there's going to be more services. And cloud services will be, we're already seeing a lot of local and nationalistic type of services or country-based services for some of that. We're seeing vertical, whether, you know, healthcare, retail, government, we're seeing. And like you said, there'll be clouds for your car, clouds for, my refrigerator could use a cloud. It's not a smart refrigerator today, but you're right. The possibilities are endless because we're making it pervasive and deployable. I mean, we spoke of the Open Data Alliance, Open Data Center Alliance. That was a couple of years ago. This was all pre-internet of things. You guys saw that coming, that wave, right? So, as Pat Gelsinger and Xintel said on theCUBE, if you're not out in front of that next wave, you're driftwood. I think that might have been an Andy Grove quote. I don't know, but we credit Pat Gelsinger. He's his paranoid, Andy Grove's his paranoid. We'll give it to either one of them. We'll give it to Pat Gelsinger. Because he said on theCUBE, Pat Gelsinger's quote, famous. Get out in front of the next wave or you're driftwood. You guys saw internet of things earlier. It wasn't called that back then. What's happened there? What have you seen now? You guys have surfed that wave and surfing it. What's going on with internet of things? We are just at the beginning. I mean, there is so much untapped potential there. And I got to tell you what I love about it is it brings Intel together. When we talk about the device across the network and communications to the heart of the data center, we are investing across an entire spectrum. And we are doing a lot of work really focusing on where the top use cases are. If we design technology and try to find a home for it, it's very difficult. When we understand the solutions that healthcare needs in terms of like the investments we've made, the partnerships with either the Michael J. Fox Foundation or OHSC on the Cancer Cloud. And there's a lot of cool wearables and IoT device stuff going on with both of those, how we incorporate data. When we understand the solution or the problem that we can solve and then to design the technology, it's better for all of us. So IoT for us has really been, it's a huge growth area. We've got a lot of new products. It's in your wheelhouse. I mean, it's connected smart devices edge of the network, right? It's silicon, but it's about everything else on top of it. And it's also good for us for partnerships because these are big, complex things to deliver. So that's awesome. So I want to ask you about security because we've been on the queue before. We talked about security. Steve Herrod, who's a former VMware CTO now and now VC is investing heavily in perimeter-less security companies. You know, you're seeing Illumio, for instance, you've got a lot of companies here. This cloud game is changing what the perimeter is. It eliminates the perimeter. So does that mean that would imply that the security has to be in closer to the data or virtualization, all the stuff at the hardware level? So what do you guys see in that? What's the strategy? What's the intel thinking on security? It's big because security is big and it's still all the data suggests. It's still the number one barrier opportunity when people are looking to explore cloud. We of course start, we have a lot of security features built in just so that you know all the way down to anything that runs on top of that silicon. For example, the TXT, which is going to secure to trust. We are also doing a lot in remote attestation and secure attestation. We need to make sure that things connect. And then of course you've heard about a lot of our, you know, if you watch the Jim Parsons commercial, right? Your face is your password and smart key and secure key. It is a chain along all of it. So security for us. And of course we have McAfee and our investments there, which we're really integrating not only into our plant strategy, but our data center strategy. So we could talk for an hour on security, but yes, it's important. And it underpins everything. But that's embedded systems kind of teases that out. You think about what embedded systems is. That's grown up. That's essentially this conference. Would you kind of tease that out for us? What is the, what's the systems on a chip mean for the folks here at the event in this ground floor of an industry? Look around small still. It is small, but it's actually really, I mean I can't believe how big this initiative is going and gone. And you've talked about the system on a chip. I mean for us, so kind of shifting just a little bit, but yesterday we launched our Intel Xeon D processor. It is the first Xeon class to SOC. So what do we mean by system on a chip? You're going to get all the performance and reliability and everything data center class of Xeon with integrated network and IO or integrated ethernet in this package for consider, which really is going to bring all that power at a density and PowerPoint that's much lower. So for us, and by the way, when we talk about integrating to your point, right, it's not just performance and gigahertz. It's security features, manageability features, virtualization features, have to get integrated smaller and smaller. And then you get cool servers like Yosemite that is going to have 96 nodes or what all stacked in there, new use cases. It's interesting how the priorities keep shifting around, but what's the number one attribute, second attribute, third attribute that you're engineering for and it clearly power and power consumption is a big one right now. After you got kind of do baseline compute, what's kind of next down the road, do you think, in terms of the shifting priority that we don't necessarily have at the top of the list right now? Well, I think for us, it's security, as you mentioned. It's absolutely making sure that from silicon to solution, security is embedded in inherent, manageability, orchestration. We're doing a lot of investments. We are a large contributor to the OpenStack community to do this as well as a lot of the commercial software packages out there to really make sure that when you have features, there's a lot of things we can monitor in the silicon that can be exposed upward to the orchestration layer. And I talked about the app controlling which resources it needs to do that. It needs to know about security policies or power or management policies or all of the different things and those attributes live down in the infrastructure. So a lot of where we're going to is making sure not only do we integrate them in the silicon but we can expose them upward. And if anybody has a chance, whether here or in the future, we've got some really cool demos to show how that works from individual silicon components to rack scale architecture to APIs that integrate to the orchestration layer. Now that is some serious tech athlete talk right there. You're talking about an operating system, orchestration, automation. These are fundamental cloud definitions that are maturing very rapidly. So now, what does that mean looking down the stack, if you will call it a stack, but to the physical hardware, there's got to be advancements there. What's going on there? You've got power and cooling concerns in the data center. Watts, I mean, this isn't density issue. It's still heat in there. So how does that change all the game? Well, I think that you just kind of pinpointed it well. When we were designing at the silicon, we were missing that. And now when you look at this as the data center is the system and come down through it and you really say what to effectively orchestrate or truly realize a software defined infrastructure vision that, which today is really kind of, the big cloud players can do it. But for the rest of the world, it's still a, you know, where they're heading to. You have to design across that. And so for us, it's why you don't just see compute network and storage silicon. You see a lot of investment in the software. And it's how those things work together that I think drive the greater result. One plus one equals three. Rage, you know, I want to get your thoughts on something. You've been in the industry for a while. Intel's been in the ground floor of the computer revolution. You go back to the history books, it's all there. We're now in the X generation of computing, new class of developers, okay? You guys attract the range of spectrum at Intel, hardware geeks and developers. You know, now you get up high in the stack, SaaS is all, you know, app developers. So the middleware battle's going on in cloud, it's passing and et cetera, et cetera. So what's the developer landscape look like? Cause here we're bridging two worlds. We're bridging hardware and software. And now it's in the data center. So that also impacts apps, mobility, big data. Can we go about in memory, in systems, in processor? The trend is inside the hardware, not so much in just memory. So what does this mean for the developer? You know, I hope what it means is more easily able to develop and create, have that compatibility to move across and get complete use and scale out of their environment. What we also want to do is, I don't want to be abstracted from the developer, right? If we were just an ingredient company, we would be. So really a lot of our paths enabling to make sure that the attributes of the infrastructure, as you said, because it is, it is about the infrastructure when you have them. And in memory database, you can't abstract away from the underlying hardware. So what we really want to do is make sure, not only to have the best ingredients, but to make sure that everything that's capable is exposed to the developer as well. And that they can easily use it. They need the tools, they need, they need it to be easy for them so they can focus on what they do best without having to worry that they're sacrificing any performance or capability by how they design. So we do, that's why we do do a lot of the open source software investments with the developer communities to maintain that. So final question before we get in the time, he'll hear, share with the folks out that your perspective and maybe Intel's or Bytes-Bursa or both, this whole replatforming trend that's going on. A lot of folks here could be partners of Intel, there could be customers out there of your customers. They're thinking about the future. And what's the mindset? What is the playbook to be successful in this new era of open source, meets doing things in the open, software-defined data center, DevOps? I mean, it's pretty phenomenal. So what's the success formula and mindset, et cetera? Well, I love what Frank said earlier today on stage, couple things, one size doesn't fit all, but also we got to break outside the box, right? The box was a physical limitation of technology, right? And it was easy to deploy. Now, what this replatform or where people are going, re-architecting the data center is really about isolating your resources, breaking them out, virtual pools of the compute network and storage, allowing them to work as a single pane across your data center where you can control and manipulate them and apply them, but having control to optimize them. And when we break outside that box and we have ability to manage our assets across, it changes the way the system vendors design, it changes the way we design, it changes the way the software is designed, and that all works together. So it's the openness and standards-based hardware combined with the openness and standards-based software allows it all to work together. Awesome, Regine, thanks for coming on theCUBE. We're here at the Open Compute Summit. Thanks for coming on. I really appreciate it. Thanks, Gillar, and with Intel, general manager of the cloud services business with Intel. Great to see you again. Again, Intel leading the charge. Great investments, you guys are doing great. We need you out there on the front line to get more faster hardware for all the developers. Faster, better, stronger. Developing in the open, that's the new model. Open Source meets hardware. It's a home group mentality. It's really awesome and live in Silicon Valley. It's theCUBE. Extracting the signal from the noise. We'll be right back after this short break.