 Okay, we're inside theCUBE. I'm John Furrier, the founder of SiliconANGLE.tv, and we are here to talk to Jim Blakely that leads the Intel Strategic Systems Integration, Architecture Engineering, and his team works on all the integration architecture with the vendors out there, and I'll solution architectures. Welcome to theCUBE. Thank you. I'm here with my co-host, Dave Vellante, and Intel made a good investment in the cloud a couple of years ago, past five, 10 years getting into the cloud. So tell us what's happening right now with your group and Intel. Everyone knows this is powering a lot of all the servers and desktops and other devices and embedded systems and a variety of applications out there. So tell us what's going on with you and your group. Well, so for us, the cloud really has two big components. One is with a lot of the big cloud service providers, the Facebooks and Amazons and Googles and in China, Baidu and Tencent and all the rest, and they've been driving a lot of architecture innovation in terms of how data centers come together, solutions come together, applications come together, which has been feeding a lot into what we do. It's been helping our business enormously over the last few years as those companies have grown, but maybe more importantly, they've been setting a pace for the rest of the industry that we see reflected in the private cloud as companies look to get to a much better, lower cost of compute on the enterprise, the traditional enterprise side. And so there's a lot of opportunity for those innovations that have been happening on the public side to make their way over and we're seeing that in spades right now. So on siliconangle.com and wikibond.org, we've been covering cloud, mobile and social for the past couple of years and we've been talking here on theCUBE at a lot of the events that we've been doing that the notion of systems is changing, right? With the cloud and mobile, the use cases are already driving a lot of change. I mean, Facebook's one big example of an application there. So what are you seeing right now about these new use cases? Obviously the architecture changes and big data is playing a role from using machine learning techniques to help do some configuration to a variety of other systems problems that are being solved architecturally through new technologies and methodology. So could you share with our audience out there what your view is of what's new now and has that compared to what was old and help us kind of clear that line of new and old? Sure, I mean, I think I'll give Google some kudos here, particularly Luis Baroso a couple of years ago wrote a great paper called The Warehouse Scale Computers in the Data Center and his basic premise was the data center's now the system and we're seeing that in spades and if you look at big data type problems and high performance computing more generally, cloud computing, it's all about turning the data center into a system. That's one component of it which is having sort of dramatic impacts on the data center overall. The second part of it is the whole client, distributed client environment as we've seen the takeoff in mobile devices and what's maybe not so visible to a lot of people is around embedded devices, machine to machine, internet of things, kinds of solutions that are coming to improve agriculture, to improve grid management and a whole host of industries looking at how they take advantage of the lower cost of computing at the edge and computing at the endpoint to build better applications, more satisfying applications for their end users. You know, innovation is a funny thing and Intel has a heart of innovation. I mean, you guys built Silicon Valley up and to what it is today and obviously going back to the roots of HP, kind of the founder of Silicon Valley. So you guys are no stranger to the innovation but our observation is that innovation always has kind of a, we've seen that movie before Factor but with a twist. What would you say is going on now relative to cloud mobile and social that you say, you know, that paradigm is kind of like that one but a little bit of a tweak. There's always never a one to one. You mentioned distributed computing. Obviously that's one. Things are connected. Systems is a systems operating system. So these are all concepts that have been researched, implemented maybe. What are you seeing? What's your view there from a geek standpoint of, yeah, this is a little bit of that and that kind of going on. Not that you're a geek. Actually, you're a super geek. I'm taking that as a compliment right there. I mean, do you know what I'm saying? Like there's always like, it's a little bit of network management kind of concept here but it's not pure network management but we've got some systems here. And then you throw big data in the mix. Yeah, so you have a confluence of a lot of things happening. Now see what people talk about is real time and faster business cycles, blah, blah, blah. Simplice of the only value propositions is where the value is but underneath it is a disruptive stuff happening. What would you share? Well, so I think the opening credits of that movie usually say at the bottom and then the cost of computing dropped by 10X. And everybody's still complained. Everybody's still complained. And I think that's, take the PC initially, take client server computing, take the growth of sort of x86 and servers. A lot of it was all driven by dramatic, architectural changes that led to dramatic reduction in the cost of how things get done in an IT organization or any kind of compute infrastructure type of situation. The other big factor usually is okay, now network bandwidth changed in a dramatic way so that now what used to make sense to do one place makes sense to do someplace else. And I think what we're seeing with what's sort of triggered off this trend is first the cost reduction on the core infrastructure, compute, virtualization, performance, the cost of storage has been dropping maybe not as dramatically as some of the compute. Networking... It's going up now. Well, volume is, the volume's going up, but the cost is... The cost is going up too because the tie floods. Yeah, so the demand goes up, right? There's a lot more need for storage. That trigger point happened. The other thing that I would say has happened too is that broadband infrastructure, at least in the developed world, has gotten much, much better and that's opened up the opportunity to do things in a very, very different way than they've been done before. You were talking earlier about the cloud service providers, the Facebooks, the Amazons taking advantage of some of these trends and innovating. They are leading the way, but actually are they? I mean, will traditional IT operations follow that blueprint or do they have to take a different path because they're managing much more complexity and many more applications? Well, I think today most of the cloud service providers, the big cloud service providers have the luxury that they're, for the most part, single application sorts of houses. Now that's changing as, you know, when Google added YouTube to their portfolio that created a whole new set of workloads that they needed to deal with that they weren't before and their infrastructure had to adapt for that. But you're right, they have a less complicated set of problems to solve because they don't have quite so much fragmentation at the application level. But that doesn't mean that the lessons that they learned and that they are, you know, can share with the rest of the industry can't be applied. And you know, security is a good example here where, you know, if you own and control the whole data center top to bottom and you have great control over everything that's in there like a Google does, security is not something you're necessarily innovating a lot within their data center environment. They're innovating at the edges, certainly. And an IT organization has to solve that problem because they have those pesky users that are inside the firewall that could, you know, potentially cause problems. Jim, a question for you on the data center, obviously the area we'd love to talk about, but in the old way, obviously data center, people think of a bunch of servers, got all the gear in there. IT guys, you know, wiring the cables up and it's a data center. What's different now that's new to the data center world that they have to pay attention to? In other words, that they have to draw up journey and path to that's not a replacement, but an extension of the data center. So, let's see the data center as a classic standpoint, lock that in. It has some innovation going on. What's the new stuff that they have to pay attention to? The architects, the CIOs, that are directly related to their world. What are the key things that you see? So I think a few technology oriented trends I'll talk to. One is clearly big data, you know, is a latest buzzword, but there's a lot of new application capabilities that are coming there. New frameworks like Hadoop that'll help drive the opportunity to solve problems they couldn't solve before in terms of data analytics and business intelligence and people really should be looking at that and how they're going to architect that into the data center. The other thing that I think is gonna be relatively a profound shift is the move to non-volatile memory within the data centers as, you know, the initial phase is a solid-state drive kind of phase where now we have a new tier of storage that can add different performance characteristics than before. As that evolves and matures, you'll see other kinds of architectures for non-volatile memory that aren't storage-based architectures that'll change the way applications need to get deployed within the data center. I think the other big thing is around automation, you know, there's actually an IDC study that says the number of administrators you need scales with the number of virtual machines you have, not the number of servers you have. So every time you add a virtual machine, somebody's got to administer everything that's inside that virtual machine. So the cost of doing that need to administer it is proportional to how many new applications you create. So that's got to get solved, right? You have to be working. That's a problem. That's a problem, right, yeah. So that immediately turns around to automation, right? Go ahead. So talk about, we've got a couple of minutes left. I want to get one last question and I know Dave might want to get a few more. What can you share with the folks out there in our audience about what they might not know about your group? I mean, obviously you work with all the people building solutions, you know, the other big guys. And Intel's a big part of that, always has been. What can you share with them that's going on now that they might not have seen and written about or might not be aware about? That's pretty exciting. Share with them your most exciting highlights. I would say, boy, that's a tough one to narrow down. There's a lot of things going on. Well, go crazy. So we're seeing, I think a shift in sort of media in graphics, in how those get deployed in applications. Historically, graphics, media hasn't been a data center sort of thing, right? It's been either, you know, IPVBX sort of solution sitting in a telecom closet or it's been a web streaming kind of application somewhere. We're seeing much more growth in those sorts of applications. If you look at VDI, we think that's just sort of the tip of the iceberg of things that are capabilities you can apply where you do some of the processing in the data center, you use some of the processing on local devices and you can build out solutions that you weren't really able to do before because you can now deploy a solution that does some of the processing in the data center, some of the processing locally and gives an optimal experience for it. Those are, that's a, I think, something we'll see emerging over the next several years. Jim, my question, and I'd love to get your technology perspective on this is, and it relates to the whole VSpec's announcement today, on the one end of the spectrum, you've got simplicity and integration. On the other, you've got this notion of choice and flexibility. Can those two worlds ever come together? Is that possible? In other words, can you codify knowledge of patterns and workloads and things like that and actually embed it into the systems and automate it in a way that I don't have to make that trade-off in the future? Or is that just always going to be a tension that users will have to adjudicate? I think it's always going to be a tension. We're, Intel's obviously a big believer in openness and standards and the idea that systems would interoperate with each other, even a best-of-breed world. We believe that that's an important proposition for end users to be able to have some of that choice. So we would probably come down more on the open side of the house in most of this and the interoperable best-of-breed. But there's a real problem that these converged infrastructures are solving, which is the complexity of integration across platforms. It's a really hard problem. It's driving a lot of cost in the industry. And the answer won't be the same for everybody. Some people will want to have, they'll be able to buy off the shelf of converged infrastructure. Some people will stay best-of-breed. Hopefully even the best-of-breed solutions will get easier to deploy going forward. But they're solving a real problem that's out there today. Excellent. All right, Jim. Well, listen, thanks very much for coming inside theCUBE. Thank you for meeting you. And thanks for watching, everyone. We'll be right back after this short break. Keep it right there. Flash memory is a hot topic these days in both the consumer and enterprise markets. And the prevalence of flash in the consumer world on our mobile phones, tablets, and laptops is beginning to drive down the cost of the technology across all markets. Forbes magazine declared flash to be 2011's technology of the year. Another say that flash is the future of the data center. In 2008, EMC CEO, Joe Tucci, predicted that the one thing that will change the storage industry more than anything else over the next 10 years is the advent of flash technology. With EMC's recent launch of the new VF Cash hardware and software solution that leverages flash technology and intelligent caching software to dramatically increase throughput and reduce latency, that prediction appears to have come true. Today we're doing the launch of VF Cash. Everybody's really excited. Obviously a little bit of press leaks over the weekend. And people looking forward to the new product, the performance capabilities it gives. We got early morning rock bands. We got a little bit of cartoon character fun. And it's good. And I think it's, I think the marketing's innovative and I think the product's innovative. Reaction's been great. You know, press analysts and bloggers in the room which has been great, but the vast audience is really online. We've had about 3,000 people online. Just a lot of good talk about how this is a compliment to EMC technology, to VMAX, to VNX, compliment to fully automated storage steering, just taking our leadership of flash just to the next level. It's really great. Flash is gonna transform the look of the data center. In the same way as the consumer experience has been transformed by Apple utilizing flash technology, I think the enterprise experience is gonna be transformed by flash. And EMC is gonna be driving it. So VF Cash is basically taking this exciting new flash technology and moving it into service in an intelligent way. We marry that flash storage closer to the CPU complex with an intelligent software layer that determines what data should sit in the server closer to that CPU. If you think about it, flash is over a thousand times faster in terms of IOP capabilities. The performance of a persistent memory at 1,000 plus X is really a game changer for many of our customers. But now if we move it close to the server, we're able to make yet another enormous jump in terms of latency, as well as bandwidth and IOPs per gigabyte. So it really is exciting for more customers from that perspective. They get the performance with intelligence, they also get that protection that they demand in their mission critical application environments. It's a continuation of the EMC flash story that we introduced in 2008, when we first introduced flash drives in the VMAX platform, 2010 we introduced fast, we introduced VNX fast cash. This is another step in that progression. The code name for VF Cash, it was Lightning and what always follows Lightning is Thunder. So we pre-announced today Thunder, which is essentially server networked flash. So something that's shareable, something that's scalable, still on the server side, but something that can provide literally millions of IOPs to multiple server applications. That's coming next. Let's say it's EMC's commitment to innovating in the core business. We challenge ourselves, we question conventional wisdom and we're not afraid to come to market with disruptive technologies that may have some impact on that core business, but if we don't do that, then we're not gonna stay ahead.