 Okay, we're back here live, wrapping up day three, or not wrapping up, winding down day three. This is theCUBE at Stratoconference. SiliconANGLE's exclusive coverage of Riley Media Stratoconference. This is where Big Data lives, putting data to work, smarter data, whatever the slogan. Big Data is big, it's huge, it's fast, and it's very relevant, both from an organic growth perspective in the marketplace, from as well as the commercial top-down, all kind of come and collided together. I'm John Furrier, the founder of SiliconANGLE. I'm joined by my co-host. I'm Dave Vellante of wikibon.org, and we have a guest who needs no introduction. Pauline Nist, a longtime friend of theCUBE. Welcome back, it's always great to see you. Thank you, thank you for having me. It's been an exciting week. You guys have been right on the edge here. It really has. Living on the edge, that's what we do. I mean theCUBE is what we do, and theCUBE is, you know, we go all day long, and it's a friendly place, very accommodating. It's a space for big ideas to grow, but there's been a lot of, more than ideas here, there's been a lot of competitiveness. Normally in big data conferences like this, Hadoop World and Hadoop Summit, it's provocative, it's intoxicating, you're talking about social science and a lot of geeky computer science stuff, but now we're getting to elbows. You know, getting a little caddy in here, and people are really starting to muscle in. EMC with Green Plum, big money player, putting the big anchor on the ground there and coming in aggressively against some of the open source players, aggressive moves, making some outrageous claims what some are saying are not relevant. Intel quietly dropping in, and also making a big crater in terms of their impact by establishing their commitment to Hadoop, which is essentially moving the entire market. Intel doesn't just wing things, you know, they do it right, and so we had Intel on earlier, and that's what we want to drill down on. Obviously, you know, we'll go into some other things, but Intel, Moore's Law, performance, security, end user experience, these are hallmarks to Intel's reputation, but you got a community here, you have an organic bottoms up growth going on. People who've made a lot of investments like Hortonworks and Cloudera, and a bunch of other open source players, a lot of partners out there, you got the top down Intel coming in with the enterprise focus, where, you know, Linux kind of grew organically, but big data's coming through the front door of the enterprise. So why not go with someone else? Why is Intel going it alone? Can you talk about the decision there and what the plans are? Are you going it alone? No, no, no, I mean, you know, you said it to me earlier, we're really in an Aboard, and we think that one of the things we can contribute here is our really strong partnerships. You know, we're not going to sell to end users. We don't have a software sales force that can do that. We have to do it with our partners and our friends, and that was the whole goal, which is we want to accelerate this. We think big data is probably one of the more revolutionary things that's happened in this business, and you know, you guys and I, we've been around at least a few years, you know? And you don't see things like this that really just take over the industry that quickly, and you know, certainly we're part of enabling that with our chips and our performance and Moore's Law and SSDs and just everything up and down the stack, but you know, I was talking to your guys earlier, we really see big data kind of becoming an adolescent. You know, we're kind of out of those gawking middle school years, and as you move into that, and you want people to take it seriously, how do you bring the right set of people to the table? Because probably the biggest commitment we've got is that yes, big data is opening up all kinds of new horizons, but for a lot of the world, it has to get integrated with what they already have, and that's one of the things we think we can help with, which is we play in the cloud space, we play in the open space, but we also play in the enterprise space, and they all got to come together for this to work. It's not, you know, sure the Facebooks and the Googles and those guys are going to drive this, you know, exactly where they want to take it, but the whole world isn't going to run on an Amazon cloud or a Google cloud, you know, they're not going to walk away from the environments they have, and it's how do you bring that together? We had Intel executive on earlier the first day, and they said, you know, for the longevity of the project, he was almost kind of hidden. I made the comment just in case someone else gets bought out, you know, Intel's basically saying, we're going to stay and maintain that commitment as a longevity. So that was kind of, he didn't really say, I said, you know, players might come and go, consolidation might happen, you know, cycles do happen, but more importantly, people tend to focus on Hadoop and certain market segments, datawares, business intelligence, but we then broadened it out and said, hey, you know, there's another world out there that Intel plays in called the Data Center, Internet of Things, and the Intelligent Edge as you're messaging now, and you know, we have devices that are just connected, and we've talked about this on theCUBE, so is that telegraphing the moves of Intel to say, hey, we want to actually, you know, I mean, it's not tele, you're doing that, so when you have the Hadoop discussions internally, is it, that's the range of conversations? Oh, it absolutely is. I mean, we are, and we have been, ever since we reorganized about two years ago and put what was our old embedded group together with the Data Center group, because embedded was really moving into intelligent devices like in vehicle entertainment and all of the retail stuff that's going on, and that really makes sure the data that stuff is throwing off. I mean, you know, you can talk about social media, and I, you know, I've been tweeting all day, I'd be the last person to say social media doesn't create a lot of data, but when you get into instrumented production lines, when you get into every car in the world, you know, seven years from now, being connected to the internet. Sensors. Sensors, you know, my joke is, of course, I'm going to expect that FedEx not only tells me it's going to be delivered that morning, they're going to come out and say to me, the guy is about to step on your front steps, you know? Go to the front door. And then there's a minority board, they're going to know what is in the package, and then there's going to be cops at your front door. Yes. You know, all kinds of predictive analytics. Don't get me there, that whole issue of privacy and security is a whole nother side of it. But I mean, that's for instance, one of the reasons that we're here with SAP as a partner because we really got involved with SAP on the HANA side, which is all real time, which is lots of instrumented production lines, which is really generating just huge amounts of data from a base that, you know, they'd really been servicing, you know, a little further back and it's now becoming totally automated and you suddenly look at, everybody worries about the variety of data, but it's also the volume, the velocity, you know, how fast is it coming in and you really want to respond to it in real time. So, Dave, what is your take on it? Because I've been kind of chipping away at everyone on Intel and EMC, I've been pretty animated about this because I really like how the big money players are moving in and you've got Microsoft clearly moving the ball down the field for Hortonworks. You've got Hortonworks owning the Linux side, which is the organic side, but Microsoft with SQL Server. I mean, no one's talking about Hortonworks. I mean, how far they've come. I mean, Hortonworks has actually accelerated and extended their lead. They've got MapR with differentiation on their side with backup and recovery and snapshotting and also disaster recovery. So, we've- Which Cloudera is now announced. They're going to do too. Cloudera's got something, so we've always said, there's a lot of beach head for everybody. You know, no need to squabble over scraps. There's plenty of room for everybody. Yeah, so my thing, John, Pauline just said you don't want to talk about security, but that was one of the most exciting parts of the announcement. So Intel open, it's good. It's like Hammond Barker says, more people doing open source, better for the world. So, I'm all for that. The security piece we've talked a lot, I've asked you, is security to do over. You know, I concluded that basically it is. And the way that Intel's doing it, the acquisition of McAfee, obviously huge statement that you're going to go help solve this problem. Can't solve it alone, but the approach that you're taking at the fundamental chip level is really what's needed here. And I think that came through loud and clear on your Hadoop distribution announcement. Actually, boy Davis was on, he said that you're playing around with maybe even adding things like cell level security to H-Base. Which is great. We've got this Rhino project. We're really looking at a whole lot of different things that we can do there. I think that's the most exciting thing that I've heard. I mean, the cell level security, we did a piece on that at Hadoop World. The big ding on that has been performance. And so, we were discussing how do you increase performance, but so who steps up to the plate? Moore's Law. Right. Yeah, so that is a good sign. Well, I mean, it's the story with encryption. I mean, I today, with what we put in the chips, don't understand why every piece of data in the world isn't encrypted. I mean, start with my healthcare records, but go to my credit card. Every time I hear about one of these credit card databases being compromised, I'm like, duh, why is this stuff not encrypted? I mean, it doesn't cost you anything. It's in the chip. That's the beauty of Moore's Law. And in the Hadoop world, we're looking at tying the encryption stuff to the compression framework and giving people a way to make it automatic. I mean, we want to make it very easy for them to have a framework that just deals with, because everybody compresses the data, you have to do that. So if you tie encryption to it and make it easy for them to turn it on and make it uniform across the framework, just make it so easy that people don't think about it. Well, you're paying up a good point. It used to be, hey, if encryption were free, everybody would do it. Well, now it is. Yes, yeah. And people are still in a little bit of denial about it. I mean, it drives me crazy. So you're in the software group now, it didn't tell, right? So talk a little bit about that group, what the objectives are, what it's all about. Well, we've actually gotten real about it. I mean, I'm one of the few people at Intel that has been a systems and a software person ever since I came in. And quite frankly, very much a little bit of a fish out of water with all of these chip heads. I mean, not that I didn't run ASIC groups and my youth and all of that. But the reality is that we finally realized that with the level of complexity in the chips and the kind of capability we could put in like this encryption, a lot of the stuff we're doing for virtualization. The work we do with the OEMs is the easy part. You get them to build a server, but if the software guys don't turn this stuff on, it's like TXT, three and a half years from when we first shipped it, we finally realized, well, duh, VMware has to support this, you know? And so what Boyd is running is what I like to call the Grand Experiment, which is it's a data center software organization. We're serious about the business and the way you can tell we're serious is that we have a software P&L. We're not just here to do an abling firmware and all the stuff that Intel is so good at, but we're here to actually find places in the market generally related to the chip because we understand we're not a middleware provider or an applications provider or a database provider, but that we can tie a lot of the capability we want to to what's in the silicon and dual reporting relationship. So it sits between Renee James, who is the head of our software group and Diane, who's out of our server organization. And the goal is, of course, we want to enable servers. We want to sell servers. That's a fundamental business. We're not walking away from that charter, but we believe that there are places in the stack that we can add a considerable amount of value. And in general, the people we're going to be selling to are our partners and people who've got products that they can embed our stuff in. I mean, we've got a data center manager product. We're not going to take on Tivoli or OpenView. You know, we can sell it to the smaller players who don't have the equivalent and hopefully get some of the bigger players to take the power management data and the stuff that's really close to the chip out of it. And the similar thing with Hadoop, which is, you know, we want to do partnerships with people who are delivering endpoint solutions because Hadoop is really an enabling technology. I mean, you know, we were just sitting over there about the fact of I struggle with how do you get, how do you get the business guys together with the Hadoop guys? Because I think there are a lot of business guys who are totally clueless about what this technology could enable in their lines of business. You know, I wrote down a quote because you guys had a Riley on earlier day and he talked about time to insight. And that's lost on the technology do's. The technology do's are worried about, you know, analytics and yeah, fees and fees and new features and you know, business continuity. How do you get an interaction going with the vertical lines of business to almost get them at the table to play to experiment with a new kind of questions they can ask. Cause I think they're all stuck in their data warehousing and their BI ruts. And you know, they don't even, you know, you sit there and you ask one query and you get an answer and you think oh crap, I could go look at this and this and this and this. And the business guys say, I know what that can't do. So I don't even bother. Right, right. Okay, so I want to ask you, I want to come back to actually the execution of your distribution. Cause I'm still a little confused on that. So you're basically going to ship chips to an OEM and in it will be your distro? Is that how it's going to work? No, no, no, we will sell the distro as a distro. So we will sign up people like we announced here SAP who will resell it and SAP will actually probably embed it with a HANA product and have a single SKU where you buy a HANA with Hadoop and it's our Hadoop. That's a menu item. That's a menu item. We are working with some providers like Cisco who will probably sell a preconfigured system with our distro there. But we'll also sell the distro standalone to people who want to get it like any open source product. So I mean it'll have a SKU at part number price discount levels the whole nine yards just like you were a typical open source scenario which is we're not selling the bits, we're selling the support contract. So we have a five by eight and we have a seven by 24 and we'll have some additive products that will do some things like some management things and other areas that we're going to be investing in over time but typical open source kind of software product. Think of us as a software company. So I'm kind of confused. I rarely get confused but I'm going to confuse them this one. Oh, I'm sure you're not confused. Dave, wait, it says I didn't get confused. Pretend I'm an idiot for a minute. It's really hard. But I'm in first grade. No, seriously, so what the business model Intel's not to make money on support contracts is it's more about optimizing silicon and building an embedded system. So, you know, does SAP need Intel support? It's kind of like you can love two people at the same time. No, the serious question is, so the business model here is very simple. You're going to get a sell support, nothing else. So with SAP. We're going to sell support. If in some areas in boys group though we do have some actual products. They're not pure open source. They're actually products that we sell. Yeah, Intel software products. Right, right. Okay, so the business model is subsidized dependent, the Hadoop, bundle that as a distro that's with a product. We wouldn't say subsidized because we think that just like Kader or Hortonworks or anyone else can deliver the product and make the money on support. We will do level two and three support for SAP. They'll do level one, you know, own the relationship with the customer. But you know, if they uncover a bug, they're going to come back for us to fix it and we'll have 24 by seven and developers in the whole nine yards. So, you're absolutely right that our number one job is to sell systems and ships and everything. And this is, you know, a way to deliver that. I mean, we've got a really great graph that I think Boyd showed in one of his slides. I don't know if he shared it with you, but our whole goal with Hadoop is we think we can accelerate adoption. And if we accelerate that in like two years. And harden it and create some innovation. But the big win is accelerating adoption. I mean, the story I love to tell people is it took Red Hat seven years to be profitable. I mean, you know, we think this could be much bigger, much better, much faster. Nobody's got the patience to say it's an open source product. It should be seven years before somebody can actually think it. But accelerating the adoption on Hadoop is very clear mandate. Right. And innovating on top of it and helping the marketplace. So, you all open source back to the community. And then, obviously, Xeons and other processors coming out and internet data center, the data center. SSE is all the stuff that you put together to build these. So, that's proprietary Intel software. So, like the semiconductor or the processor, you encapsulate the complexity. That's the goal of Intel, right? To kind of create some innovation around hardening so the developers and... Right, and, well, not even just hardening. I mean, the problem now is you have three billion devices on the chip. You can do a lot with three billion transistors, you know? And so, you get into these conversations of, okay, wow, we really did a great job with encryption. I mean, I think we've got these AVX instructions, which are really vector instructions. Well, it turns out that what's a column store? It's a vector. So, we've got like every database software guy with column stores now using these HPC instructions that we've put in the chips that we're going to be in our third generation of coming up. So, you know, it's features like that where you say to them, if you turn this on, you're going to make this four times faster. There's no doubt Intel has muscle. So, when I first saw the announcement, I didn't really go to the place where everyone went, which was, oh, Intel's coming in to muscle. Unlike EMC is more aggressive. I think maybe you got a little bit of a negative halo effect there, but there are negative effect from EMC, but I didn't go there with Intel. I first thing I thought of, wow, they're going to bring some IP and bring new range to a data center, stuff that we've talked about, you know, probes in the data center, bringing that software together using big data, log files, a lot of things. So, to me, that is much more of an interesting conversation. So, what are some of the things around this intelligent edge that are software related that you think is lifted by big data? You know, I will just tell you that we first got involved in this business in China because we were getting companies, there's an interesting thing about China. They have no infrastructure to walk away from, so they don't have teradata's in the back room, you know, they don't have all of this previously installed stuff, so as they look to solve problems, I'm not taking a slab of teradata, but I'm just saying that, you know, whether it's Oracle or whether it's IBM or whatever, it's kind of like why they got into wireless. They didn't have copper. I mean, you didn't have any copper. If you didn't have any copper, you didn't even have wireless telephones, right? I mean, how do you make it work? So, you know, you're starting from scratch and they're looking at a lot of these smart city programs where they're trying to instrument these new cities they're building and they're just creating volumes of data and a lot of the early work we did with the intelligent system stuff we did over there. And I know those countries, the bandwidth has been phenomenal, so the flow of connectivity has enabled a lot of machine data, user data, social data, much more than the US, you get the telcos here kind of still last miles always been at. Yeah, we were talking about that earlier, that the bandwidth is going to be one of the limitations here. Okay, so give us an update on what you're working on right now, because obviously you're always, even I was saying the other night, like, you know, you always got something going on that's new, so what's new now? Obviously your software and systems background is really lending a nice expertise for Intel right now where they are, honestly, you know, with the embedded stuff coming together and the intelligent systems. So what are you working on? What's your focus? Well, my focus has really been a lot of the partnerships for these products, because I've got, as you guys know, I've got a long, long history in the enterprise IT space. So, you know, I've known the SAPs and the SASs and all of these players for years. And so now it's really been helping void the rest of the team, put together some of these reseller arrangements, go in and really kind of, you know, we've got lots of agreements with all of these guys to do NDAs and everything else. We've never actually dealt with them as a software partner where, you know, we're talking to them about how can we not just get you to resell our product, but what we really want you to do is work together with us from an engineering point of view to make the products better. So I mean, we know that SAP, for instance, is not going to turn down a customer that wants to run Hortonworks or wants to run Cloudera. What we want is them to do work with our distro to integrate it better into HANA, to have it be a better solution that is more closely coupled, that takes advantage of the technology agreements we've had for years with them, because they're already doing things to take advantage of our processor technology and our SSD technology. So we think that we can take those relationships to the next level, and that's what I'm doing with my team. Awesome, Pauline, it's always great. I'll give you the final word here. Explain to the folks out there, just your take of Strata and the big data world and what's happening and what people can expect over the next 12 months or so. Well, I think it's fast moving. I mean, you can't have been here this week and not be energized by the fact that there are people all over the place who want to do stuff. I mean, I love the fact that you could pick up and either hire me or we're hiring ribbon to put on the bottom of your bag as you walked in the door. This is one of the areas in tech that's exploding in growth. Yeah, it is, it is, exactly, exactly. And so that's exciting, but like you, I do believe that we're going to see, we have to see over the next couple of years consolidation. I mean, there's just lots of people running really fast trying to do really innovative things and that's what Silicon Valley is great for. I mean, it's great to have a show like this here in the Valley, but I also think we're going to continue to see changes in evolution as we go forward. Evolution, consolidation can be a good thing and then it's growing up adolescence now, soon to be teenager and then young adult and fast. So it's like that scene in Star Wars with the clones that gets grown really fast. And I was like, this is going faster than anyone's ever predicted. I guess it is, it is. The commercial interests now are front and center, yet the organic open sources are still exploding. So great stuff falling in this inside the cube. This is Silicon Angles coverage of Strada, O'Reilly's Strada conference. We'll be right back with our next guest after this break.