 Hi, I'm John Furrier with SiliconANGLE.com. We're reporting at Intel IDF 2012. I'm here with the general manager, David Tui of Storage Group at Intel. Is that the right title? Storage? That's right. Storage division at Intel. That's right. You're the general manager. You're the head honcho. So you have the business line objective of storage. So first, what's happening at IDF here? Obviously a developer forum with a lot of alpha geeks setting the agenda for the next few years. The normal stuff. You see a lot of graphics, a lot of great demos, intelligence systems, stuff in Intel's wheelhouse processors. So tell us what's going on here for developers and what's going on in storage and how's that changing? Yeah, so in general, this is our big Intel developer forum IDF and it's our show to reach out to the community. As a chip supplier, a building block supplier, we completely depend on the ecosystem to be able to deliver our value. So we're fans of everybody that does software, hardware and solutions, right? And so for us in the data center, we've been focusing on cloud. It's a big topic, of course. Actually, there's even a bigger topic lately called big data and in high performance computing. All three of these things are actually interplaying with each other and they're affecting next generation data centers. So we're talking about a series of technologies and solutions that we're doing. We're doing with ecosystem partners and we're innovating on in the future because we do a ton of R&D and Intel, we spent 13% of our budget on R&D for three to five years south. That's what we're talking about here at IDF. A lot of your R&D is kind of coming home now to roast with some good solutions. You had a great investment with cloud. That paid off pretty well. Invented systems, you guys had been doing that stuff for years. Now the kind of world turns your way now with big data analytics and the data center becoming more software driven. So share with the folks out there, what does this notion of software data center mean? Because you guys move from being a component player to more of a technology provider in the data center. It kind of changes the equation a little bit from what Intel kind of was doing. Now you got a big presence in the data center. What is this new error that you guys are forging into with this whole software notion? Yeah, it's actually very fascinating. Of course, our strategy is always to be a building block supplier, but we do have a significant amount of software building blocks too. What's going on in the data center is pretty radical. In some ways, there's there's a big push towards server based virtualization over the last few years. And that's actually been very successful, right? You can find many companies have 75% of their servers are virtualized. But it's not been true on the networking or the storage, those have still been dedicated embedded appliances that are that are provisioned and managed separately. And so what you're beginning to see now, and this is that the very big cusp of coming into existence is this idea of software defined networking and storage virtualization and storage becoming part of that infrastructure. And so, you know, it's going to take some time, but that that's actually a very fascinating way to build next generation data centers. And that's that's really cool. In traditional enterprises, actually, there's a whole movement of foot with unstructured data, where they're going from, you know, traditional data they grew most of was frankly, their ERP systems or their Oracle database SAP database. Now with the advent of, you know, the internet and multimedia, unstructured data, video media files, that's really becoming the big workload. And so they're beginning to ploy a new storage architecture called scale out storage. And we we talk about we call it intelligent scale out storage because they're using not only different methodology to build the storage is actually much more modular. But they're actually using a lot of very powerful math algorithms to save tons of disk space, right? So as these disc grows, things like deduplication, thin provisioning compression, these technologies are saving a lot of infrastructure for IT organizations as they build out this this new data center architecture. We're just talking with cloud air one of your partners who have obviously been the first commercial venture back company on with Hadoop. And one of the things we're talking about was the more data beats algorithm kind of philosophy is that people are acquiring data very, very rapidly at a mass scale right now, which affects your business, which is storage. But yet the paradigm is now moving to near real time latency. So that combined with say, solid state disk tries or flash has kind of changed the equation in the architecture. Can you talk a little bit for the folks out there who aren't inside the ropes around inside the industry around the notion of big data, acquiring lots of piles of data with the ability to bring it back really fast in real time. And also with solid state, essentially, non spinning disk, which is like having a lot of memory. How is that affecting the developers? And what's that going to lead for solutions? Yeah, I think you're hitting on a very important thing. So there's a real actual bottleneck to being able to scale to these quote big data workloads, which are defined by a combination of just huge volume, a lot of variety, very different types of data that they've never been able to look at before, and a large velocity, you know, this idea of being able to look at it real time. And so the classic disk is a wonderful technology, but it's not getting fast. It's getting bigger, right, but it's not getting faster. And so it's actually one of the bigger bottlenecks to be able to get scale in a data center. So nonvolta memory is coming in to fill that gap, right? And it's being done in the form of traditional solid state drives, which people are now plugging in the back of the servers. It's a very big part of a Hadoop workload. That's a very disc intensive activity. So several of the large public data centers are very heavily prevalent using solid state drives to take care of the Hadoop type map produced search workloads. But it's also affecting the entire tiering structure of storage. So now you have a storage tier appearing in the server, a storage tier appearing in the network and storage tier appearing in the in the capacity store. All of that is being affected by nonvolta memory is actually starting to re architect the whole data flow through the data center. It's very exciting technology. And it's really, you know, the the nonvolta memory technologies really what's enabling that to happen. You know, when I hear re architecting the architecture, that means opportunity for entrepreneurs and big companies like Intel. How big is this? I mean, and try to put your finger on is how massive is this disruption solid states causing in that regard? Well, it's already you've seen small startups become multi billion dollar market cap companies, right? You know, so I think I think that nonvolta memory over the next five to 10 years is really going to, you know, fundamentally change how we do data management and data flow. You know, there's been a big issue that the you need to have resiliency and have these really, really wonderful, but very expensive, you know, solutions to be able to store the data over time reliably. But it's also on its own separate network, they call them storage area networks for a reason. It's a separate complete appliance, right with multiple boxes ability to deliver it. You know, many companies like emcee with fast and end up are doing these different technologies now where they're reaching from the storage all the way into the server and deploying value add capabilities using nonvolta memory in the server. So they're distributing that storage, which makes it much more local to the computer app and much much faster. And so that's just going to continue to evolve. And you know, in revolutionized to also from a different vector. So I think that's a key technology. What goes with that, though, is networking, right? Because you got kind of a data execution problem, you've got data flow, and you got a data capacity problem when you look at these these data centers. So people are really starting to embrace 10 gig, for example, is really being driven hard by the cloud and IP DC usage models. And that technology is going to go to 40 gig pretty fast, much faster, I believe that we've seen in the past. How has a final question for you know, you got to go thank you for your time here at the software dot intel dot com developer zone here at IDF in San Francisco. Final question, how has this big data change over, if you will, or the expansion expansion in this new area, as well as solid state change your business at Intel in terms of how you're organized and how you go into market and your and your relationship with your partners. Well, it's actually been a wonderful thing for us, frankly, because all of this new technology requires intelligence. And, you know, the core element of our roadmap is CPU technology, which is basically delivering intelligence and enabling intelligence, right? So what we see, for example, some of our highest performing processors that we make are being used in these deduplication appliances, right? And, you know, compression is a huge demanding workload. Nonvolta memory coming in and providing a caching tier is a huge demanding workload. So we're we're actually very excited about it, you know, because it helps us deliver more value. What goes with that then is the power, right? So what we see is some of the earliest adopters, some of the first people to take servers off the line for our customers are the public data centers. They need that power performance. They get an instantaneous win on their TCO models by taking Moore's law and instantly deploying in their data center. So we get we get a great power story, we get new technologies that require intelligence. And so we have every day we have a new company that comes and talks to us as being funded in storage by venture capitalists. And they want to use our technology, which we we sure appreciate, you know, we like that a lot. We have a great supplier always pushing the envelope enabling and enriching lives around the world. Intel David Tuey, I'm John Furrier with SiliconANGLE.com. We're reporting here in San Francisco at Intel Developer Forum 2012 in the developer zone software.intel.com. We'll be right back with our next guest.