 Okay, we're back at Oracle OpenWorld, and I'm excited to bring you live coverage from Oracle OpenWorld Live in San Francisco, California. I'm John Furrier, founder and CEO of SiliconAngle.com, and I'm here with my co-host. I'm Dave Vellante of Wikibon.org, and we have a great guest, Pauline Nist of Intel's high performance server group. She's the general manager of that mission critical operation, so welcome. Thank you. Pauline, we talked at your Intel developer forum, which happened a few weeks ago here in San Francisco. We didn't have the cube there, but if we could have gotten a gorilla sneak in like we did at Oracle, we would have, but it's always hard to get that access. But we had a chat at one of your press events where we were talking about, you know, big data and the data centers. It's really impressed with the caliber that Intel has, and your background in particular, you've been in the industry in a variety of different capacities. For too many years, I was hired as a child. You've seen, you've seen, and you're really technical, and you've seen your share of paradigm shifts. And, you know, Intel's has always been a dominant force. We all know Intel as the brains behind the PC revolution client server, and now with mobile and power, and all that stuff in the data center, a lot of things happening. So tell us just quickly what's going on with Intel in the data center, and some of the key trends that you're seeing that are really the most exciting for you. Well, I think we're really excited about the acceptance that our high performance zeons have gotten. You know, we announced Nehalem about a year and a half ago, your girl last February, eight socket systems for the first time, and now you're seeing partners like Oracle with the ExoBoxes, Exologic, Exadata, Exolytics now. We all say the X stands for Xeon. Wow, nice, that's a good angle. That's a good trend connect. But, more importantly, what I actually see happening that we've fostered is almost this revolution in how people are looking at dealing with data. You know, whether it's OLTP or whether it's analytics, we're putting a lot of performance out there, a lot of capability out there. You put our chips together with SSDs, with the kind of networking today. You're seeing people doing mission critical software as a service, but more importantly, you're seeing innovation. That's an amazing thing to me, whether it's Hadoop in the open source world, whether it's what Oracle is doing on Exadata. We think one of the most fascinating things about the Exadata box is it's got Xeons on the storage side and Xeons on the server side, and it's actually changing where the computing is taking place. They like to talk about that Infiniband pipeline, but if you screen the data, if you filter the data on the storage side, you don't need to put it across on the server side if you're not going to need it. And so, to me, it's the innovation across the board, which is whether it's open source, whether it's new suppliers, I mean, you got EMC buying Green Plum. Who would have said that was going to happen? You got IBM buying the TISA. You've just got a lot of innovation happening and manipulation of data. So Pauline, you have a very strong background in mission-critical computing. Your tandem, right? Yep. Digital. Digital. So you've seen quite an evolution, and there was a time when it was all about scale up, that's really the way to go, and Intel has really changed that along with some of your customers. Talk a little bit about what's happening in that whole space. So, obviously, a lot of excitement about scale out, but we could all admit that the industry's been talking about scale out versus scale up for about six years now, and- Or 16. Or 16. My attitude is you're obviously going to see both. I mean, there are still some workloads, large Oracle, OLTP databases that are going to want scale up, a lot of innovation in the scale out space. I don't think it has to be a choice. I mean, Exadata at one level is an example of that. It's two eight ways. Could it be three, four ways? Could it be two 16 ways? It really would depend on your data. I think the scale out space is driving a lot of the innovation, because people are putting together these cluster configurations of servers and doing a lot of analytics. As we all know, I mean, the interesting thing about analytics is it's read-only, so it's really easy to scale that out. The minute that you've got to get involved in sequential transactions, life gets a little more complicated. Well, then you're right. Scale out seems to be where all the innovation is, but your competitors have had to make choices. I mean, they seem to be in the box of, okay, we're going to go after that. I don't know what the percentage of the market is, but it's the smallest part, which is the- Well, it's the smallest part from a volume space, but as we like to say, there's still $15 billion worth of server revenue out there that's in risk and mainframes, and that's a lot of revenue. It may only be 2% of the volume, but it's a lot of revenue. So you're going to go after that next, right? Yeah, but we want to bring the power of open to all of it, I think is the way I would say it. We used, yesterday, John and I were talking with David Fleuer, an analyst at Wikibon, and we were using a horse racing analogy. We had, I don't know if you follow horse racing, but we had secretariat winning by like 31 lengths, that, well, that was Intel. We had a couple of the competition down, a little behind and focusing on some of those niche, $15 billion niche markets, but it's just amazing to see what Intel has done. Now, people used to look at Intel and say, well, it won't be in the server space, it's really a PC, so, you know, a processor. What do you think about a company like ARM and the mobile? I mean, will that ever contend for, you know, mainstream data center technologies? We are, I will say, not protectionist about anything. I mean, we are looking at bringing out- The paranoid. Well, you know, we're looking at atom servers, you know, if there's a web hosting environment that is going to benefit from a lot of smaller core servers, particularly if you have a workload that's very homogeneous and can scale across that space, I don't see Oracle running a database on it, but on the flip side, I could see that, you know, there are web hosting environments where that might be a very competitive offering. And so, you know, we're actually working with those guys to characterize what workloads work best on smaller cores versus bigger cores. And I know the Google guys actually did a paper called Wimpy versus Brawny to kind of deal with that workload characterization issue. A super low power. Yeah, exactly, exactly. So in the world of Oracle, obviously you make those points with OLTP, et cetera, et cetera. But now, Mark Benioff's kind of un-keynote that he did that we covered this morning at the Regis, he talked about Facebook, he talked about these new environments, and we were talking with Jim Lundy, ex-gartner analyst who now runs the Argonne Research talking about how socials be more pervasive and that app firewalls are becoming more of an issue in these enterprises, et cetera. But this brings the trend around high performance cloud computing in the sense that Facebook is an indicator of where companies may go in terms of building out commodity hardware at a large scale. What's that market look like to Intel? And also, you guys want to sell more processors, you have great power management, all those things that are in your wheelhouse. But like, as that evolves, will people look more like Facebook? Will there be more purpose built with open source and Intel gear where Oracle's obviously got their purpose built machine? So what's your view, just from an industry standpoint, around the big movement towards these open systems? Well, I'm going to go on record as saying that I think that for Enterprise, it's going to be a journey to the cloud. And there are good things that people are looking to steal already. I see a lot of people who maybe wouldn't be characterized as cloud the Enterprise space, but are software as a service and have standardized their configurations. You know, it's the Southwest model of all your planes are 737s. It's a lot easier to train your administrators and all of your people to take care of your configurations if you've only got one configuration. I though, I'm going to go on record as being a little out there because I think that when you get to Enterprise and actual data where it's not just analytics, it's actually transactions, whether they're financial or in our case, planning our supply chain, what chips we're going to build, how we're going to get them through the supply chain. I think you're still going to see some data optimized configurations that look different than the web hosting configurations. I mean, they all may be built out of standard parts, but you know, you may want bigger memory, you may want lots of SSDs, you may want processors that are higher performance and can support those big memory configurations as opposed to what you might want at the web tier around that, which is app serving. And I just think it's going to make sense to do that because you're going to want the most effective data management that you can get and the most effective application server. And even with Oracle, you're seeing some of that. I mean, I think it's the difference between an exologic box and an exadata box. They're all built with Xeons, but they're tuned differently, they're configured differently. And you know, whether you do it with an Oracle product or whether you do it yourself, you know, assembling systems that you buy from the OEMs, you're still going to want to get the best performance you can out of that package. So you're suggesting then, correct me if I'm wrong, but oversimplifying it, the general purpose computing is going away. To use case-driven purpose? And I think the Facebooks and those guys are examples of that. It's a use case, but it's also a tiered environment. You know, they've got the load balancing stuff out there right on the edge of the web. They've got an app tier in there. Most of them don't have to deal with real transactional data. I mean, you know, if you're on travel velocity, I guarantee you when you give them that credit card, you're not sitting on the same machine you were searching for flights on. And you know, that's my theory, which is, yes, it's tiers and they're all going to be open, but you're still going to want to optimize, you know, and maybe those atoms are at the bottom tier where you're right at the edge of the web. You know, I don't know, it's going to take workload analysis to figure out what the right answer is. Does that change the workforce configuration of deployment for the labor side of it in IT? You know, like I said, they're all, you had 747s. Now you have 747s and you got small planes and you got different kind of runways, if you will, to work with. So are people ready for that? And how do you see that transforming? Well, I don't think people are ready for it. I mean, my joke about UNIX used to be, I mean, I've been to the UNIX conferences, you're going to pry UNIX out of their cold, dead hands. Yeah. But, why? We know those days. Yeah. I thought there was snake oil back when you were here. Right, right. But I think they'll get efficiencies. I mean, I think there's still going to be specializations. I still don't think that, you know, the guy that's doing the web hosting is the database expert. But the companies are trying to automate that. I think everybody's striving for more automation, whether it's in provisioning or management or performance tuning. And as you get more standardization in there with open source tools, you know, Hadoop is a great example right now. Everybody, it's the hot word. Everybody thinks it's really sexy. It's a whole pile of software and none of it's been particularly optimally tuned. And when you go out and you look at these Hadoop machines, people are running. We've seen some of them with like 14% utilization. Now, you know, as a hardware sales person, I love 14% utilization, but it's not going to stay that way. It's going to get better. And I think that you'll still find areas of expertise and what's being taken from the hosting sites is that standardized and configurations train people, you know, that everything, you know, looks the same across that space that every machine doesn't have to be configured differently. In fact, I think blades is the first step toward that. You know, standardized and a blade configuration, that's just a different version of a standard configuration that makes it easier for people to provision, manage, you know, turn on, change the usage models on in terms of performance and... So that extra compute power that's laying around, Amaral Adal who's the founder of Cloudera, we've talked on theCUBE and he's highlighted these. Like, you know, there's an abundance of compute. Compute's not the problem. Storage is the problem. So now this comes back down to the storage. How are people thinking about optimizing the big data architecture to maximize compute with the storage? Are you seeing anything pop out? I think the revolution that's happening now is just the first wave of SSDs. I mean, people have done the quick, dirty, cheap thing, which is pull the hard drive out, throw the SSD in, get rid of the latency, get rid of the rotation, move that data to the processors faster. But that's really just the quick and dirty way. I mean, we've still got that horrible storage stack in there. I think what you're going to see in the next round of products is SSDs we more directly connected in. And then I think you have to look longer term at the whole evolution of non-volatile memory. And do we end up at some point with two level memories and things like that that keep those compute cycles moving? I've always said about the CPUs, the sooner or later you've got to worry about feeding the bits into them. And you know, probably the other thing that's out there is optical photovoltaics right onto the chip at some point in the next five years. Because there's a limit to how many wires you can attach to that silicon. Pauline, I just got to say, you're cube gold for us. We love your cube material. You're the first person you ever said photovoltaics. Yes, yes, yes, okay, let's continue. No, this is fantastic, we're going to continue. Let's cut it, let's just cut Larry else's keynote out. You go ask everybody in this business, and some weren't a lab somewhere, I don't know if it was HP labs or IBM labs or the telecoms, people were playing with optical. And the horrible storage stack is also what's happening. So optical will rear its head in the R&D, hopefully we'll be around, and hopefully the cube will be still around then. Big data. You're going to have chips and planet in your head and your thoughts are going to go directly out into the internet. Speed of thought, that's what Larry said. Can bosses just go that fast as John Sprang? Larry said the big data analytics at the speed of thought. So big data. We talked about big data privately at the IDF. That's how I met you, and didn't even know who you were at the time, but you were into the big data thing that we were doing. What is your view on big data? Just at a big picture, just the future, what that can bring to us. You know, obviously NetApp has the money ball, the guy who wrote, or the Billy Bean, who is the subject of money ball. It's going to show that big data could be for anybody and that you can transform. Right. So what's your view? I'm going to give you one little quick sidelight comment. They had to get Brad Pitt, because Billy Bean's actually cuter than Brad Pitt. He is. And the line was around the corner, that's for sure. Photo opportunities. So I do think we're going to go through probably this explosion of data before we get our hands around how to manage it. I mean, when you just look at the devices that it's being generated with and that the automakers are looking at putting the internet into every car, you know, you got digital signage now that is watching who walks by and trying to flash you deals on your cell phone at the same time. The term I've heard in the last week that I was fascinated by is to talk about big data and to talk about exhaust. In other words, you get it all. How much of it do you keep and how much of it just goes out the back? And that I think is something we're learning because there's just a lot of it out there and how much of it is going to turn into information versus how much of it is going to be stuff that in the end game, you don't need to keep or you don't want to keep. If you were tasked, let's just say hypothetical I'll see your big time executive at Intel. Let's just say hypothetically, you were a big data. You were in charge. You were tasked by the CEO or CIO or COO. Pauline, we're a Fortune 100 company. Figure out our big data strategy. What should we do? What would you recommend that a company, what steps do they take? What would you suggest in terms of how to approach big data? Well, I think it probably, this is an interesting show to talk about it at because you have a different focus on big data if you're in the consumer business versus if you're on the server side. Because in the consumer business, you just want it. You want to get your hands on more information because it's the key to selling things to people. If you know, even if it's anonymous, if you know the demographics, if you know what attracts people, you've got vendors right now with digital signs that have got signs that can track your eye movement and say what on the sign is causing you to look at it and measure, was this part of the ad more effective than that part of the ad? Because they want to get your consumer dollars. The real question is when that comes back into the corporate world, what do you do with it and how do you turn it around into improving your product development? I mean, even in our case, more and better consumer data about what usage types people have for what devices then feeds back into your product development cycle and your business cycle of how do we do better solutions. You've got us totally turned around, focused on ultra books now, which is give you something that's as thin as the iPad but with a real keyboard. What are you learning from how people use these things? Because I laugh when I go to conferences and see what percentage of iPad people unpack keyboards that they're carrying along. Because I mean, what's wrong with this picture, right? Exactly. Okay, so back to another topic that's hot to us at SiliconANGLE, it's mobility. I'll say cloud is a big power base for powering the edge of the network, a device that needs low power, good battery life, again, all the things you guys work on. How does these data center architectures shifting in the cloud where mobile will be really powerful? I mean, it's already powerful now, but we saw iPhone announced the two radios. It's going to actually, that's an Intel opportunity in a way. It's got RF there and it's got chip sets. And what's the vision for how mobile and data center work together? Because the data center is whether it's a service provider and or an enterprise, the enterprises are turning into service providers and away with consumerization of IT. So can you share your vision and how you see that all coming together? Well, we talk about the compute continuum and it's that the servers really have to be smarter about what the device is. I mean, my boss uses a couple of great examples. You wouldn't want to start a backup on a laptop. If you were smart enough to know that the battery wasn't going to last more than 20 minutes, you know, because why start shipping all that data, you know, up to the cloud to back it up if you're not going to have the battery life. The same thing about how smart the devices are, we get into a lot of conversations with cloud providers or a lot of them want to build heavy duty graphics into the server because they want to assume that the device they're servicing has no graphics capability. Well, that has a lot of heat and cost on the server end versus if you had different server farm configurations and you were smart enough to know what kind of graphics you had. If the device has got the graphics, don't do it on the server side, ship it down and let the device do it and minimize the power and the cooling and the configuration you've got on the server side. And then separately, if you're smart enough about the devices, you can know when somebody's moving from one venue to another. In other words, when you walk into your home, you would really like to know, okay, he's carrying a cell, but he's going to walk into a place where he's got internet connection in the TV and internet connection in the PC. How do I transfer the information that you're looking at here onto the wall? I mean, I have to tell you, the cutest compute continuum commercial I ever saw, it wasn't actually ours. I don't know if you saw the ones the NFL ran last year where the guys in his living room and the game's on a big screen, I'm an NFL fan, you can tell that. He folds it up, takes the game with him to his car on the cell phone, then goes off to a coffee shop and has it transmitted to an iPad, goes through a couple more gyrations, comes back into the house, throws it on the wall and he hasn't missed a beat, the game's up on the wall again and you want to do that with movies, you want to do that with email, why not be able to take it with you regardless of what the device is? Paulie, okay, I wanted to talk about, we talked about Cloud before, we had Pat Gelsinger on theCUBE a number of times, 30 year Intel exec, we asked Pat in May, Pat, is cloud security a do-over? And he thought about it for me, he says, you know, yeah, I think so. And I think it starts with you guys. It starts exactly why we've been talking about trusted execution and why we've been doing the Open Data Center Alliance, which is at some point, whether it's virtualization, whether it's cloud, whether it's the data, as it moves around, you've got to deal with security from all the dimensions. I mean, our latest chips have encryption instructions, so you can literally encrypt everything without it costing you any performance. I mean, we've made encryption nine times faster. Oracle issued a press release right after we released the parts. So we see that things are going to have to be encrypted, but we also see that you're going to need to control virtual machines. How do you know that it's a trusted virtual machine and keep people from setting up VMs that shouldn't be there of an attempt to access your data? And then you get into just protected data. I mean, we're just talking to a healthcare guy. You certainly don't want anything in terms of your electronic medical records. It isn't access controlled, encrypted, the whole nine yards. And we aren't thinking about that today. I mean, Facebook's a prime example of that. Anybody can put anything up there. They can put a picture of you doing whatever tag you show it to the world. I love the Las Vegas commercial where the girl gets shunned and at the end of the commercial, she says, okay, I'll never put anything on another picture up on the net again. And that's the problem right now. We're in a total Wild Wild West. And as enterprise starts to adopt these cloud things, you're going to see, I think, some law and order come into the equation. Are you a Patriots fan? Patriots second, Steelers first. Oh, okay. That's cool. Let's speak volumes about who you are. Okay. And we love the Steelers, AFC. How can you like the Patriots second if you're a Steelers fan? Okay, I got it. Well, I've been Boston for 20 years. After suffering through 20 years of them losing, I get to occasionally celebrate them winning. Okay. Well, we love sports analogies here. The ESPN of tech, as we call it. Love to have you Pauline Nistas with Intel general manager of the data center group who's got a diverse background and has a great career, has seen a lot of the evolutions of high performance computing from data center to mobile. And we're living the age now and where the software's the key part of it. So my question to you is, kind of just putting your guru hat on. A lot of these challenges you're talking about is software related. And all the rage is software innovation, software switches now, things like open flow at the network level is kind of being talked about and software in the phones and you mentioned all this. How does all that come together with Intel? A very software specific company, but at the chip level. How are you guys developing your software strategy to create more intelligence, as you mentioned, to get those scenarios down? And how does that relate to the developers out there who do like, you know, Haruku and Ruby on Rails and C and all that stuff? Well, I think one of the big things that's changed about us in the last few years is it used to be when we were worrying about what we did for the next generation chip, we'd go talk to our OEM partners, the guys who did hardware. And what we've now realized is you've got so much intelligence on the part that that's interesting. You know, you got to worry about connectors and you know, how you get the bits on and off and you know what the configurations are going to look like. But it's really being exploited today by the software people. I mean, when you want to talk about do I need encryption in the chip? What we talked to Oracle about is how can we make your decompression faster? The compression end isn't the issue. It's when you want that data back, you want it decompressed instantly. How can we put instructions in that will help with that? How can we speed that performance up? What other things would you like to see? Whether it's from the web point of view, you know, with the new tools like Ruby and the LAMP stack and everything, or is it on the database side for tools that will help you manipulate the smartness of the data better? To even things like Hadoop, which have a totally different structure. So we find ourselves spending more and more time talking to software partners and getting ideas from them about what goes into our next generation parts. But also, I mean, it's one of the reasons we're at conferences like this. It's one of the reasons that we do IDF because you've got to do everything from the client side up through the server. And we are looking for different things on the server side than the client guys are obviously. There's some about HP's tablet and Amazon's Kindle Fire. So some sales figures were leaked on the web when we covered on siliconangle.com where the tablet sales for Kindle Fire were surpassing the iPad. iPad also is out there, it's a genius product. But expensive. HP had the touchpad and then when they had that big snafu dropped the price to $99 and boom, off the shelves. So this seems to be a price point. Yesterday, Apple announced the iPhone 4S, seven times performance on the graphics processor, two times performance on the processor. So you got the performance Moore's law thing going on at the edge. I mean, so there's issues there that are being worked on and getting better. The question is price point. There seems to be a nice price point in the consumer market for a device like the iPad and the iPhone at around $60, $99. So how does it come in like HP get there? Because they're struggling and other companies that are in that hardware business. I'll give you one comment because I heard you guys talking about the iPhone before. You know, new, better, faster, more wonderful is always great. Did you notice the iPhone 3? Now you can get free with a contract. And I mean, I think that's going to open up a whole new tier of iPhone users at the lower end who are maybe looking at other things. But you're absolutely right. I mean, what we were amused with about about poor HP in their tablet was that at $99, an item that I think had a bomb cost over $200 was really flying off the shelf. So it was currently one of those opportunities. That's such a new item. Fix that. Fix that, right, right. The thing about the bomb, by the way, is bill of material. What the actual cost of the parts is. When the parts cost twice what you're selling it at. That is not a good thing. That's not good for the company. But getting back to the Amazon, I think the piece you're missing there is it's not just the device, it's the infrastructure around it. And I think what made Apple so popular was the App Store and what is making the Kindle Fire so popular is, I mean, I'm a Kindle user, the old black and white version, just because I love the ability to just, you know, double click on 12 books and they're instantly on my Kindle, you know, 30 seconds later, how can you beat that for a buying experience kind of thing? And I think that's what Amazon's bringing to the table. And that may be a hint of, you know, for HP, it's not just a price point issue. It's a what's the infrastructure behind it. It's the Gillette blades, you know. You get to pay for the blades and you get the thing for spree. Okay. So do you see that more? Open up a checking account. Here's your tablet kind of thing. Or is that cross-pollution? Oh, I don't know. I've been watching the banks. They're making me pay to get my money out. They ain't going to be giving me tablets. Okay, so unless you keep a minimum balance. I have a $15,000, right? Okay, so you're an industry legend. You're around and you have all this perspective. That's a dangerous word. So Steve Jobs stepped down. This is their first big keynote they did without Steve Jobs. And I'll see it wasn't that big of an announcement. It might be where they come out of retirement. But we have Larry Ellison, he's still kicking. Oh yeah. He's coming out and he's doing his thing. He said he's a little bit tired from his son's wedding the night before. So he also flew in off some personal downtime and kind of bombed this keynote. The teleprompter failed. It was a total disaster. I've been through a teleprompter failure. That's not good. It was a total disaster for Larry. You age about 13 years and I think you lose one of your nine lives. He was probably sweating his shirt. He was probably soaking wet. Okay, so he's out there, he fails. Mark Peneov gets kicked out of the conference. His keynote is rejected by Larry in Oracle. He's been summarily dismissed. So Mark Peneov's out. He does the gorilla keynote. Larry's going to come on at 2.30. What do you think about all this? I mean, this is like the slug fight, the egos. I mean, well, the egos, the egos of master. The egos have always been here. Larry, I live here in the valley, you know? And Larry still gives the best sound bite. You guys got to admit that. Absolutely. Great for us, great for ratings. We miss Scott McNally, you know? We don't have nearly as many colorful CEOs as we used to have out there. I'm looking for Larry to redeem himself at 3.30 this afternoon. I think he's got an extraordinarily high bar. Is that one it is 3.30? Yeah, it's 3.30 this afternoon, so we're all going to be there watching. I thought 2.30, but maybe it's 3.30. Today's Benioff thing, he's going to come out. It starts at 2.30, but I think he doesn't go on till 3.30. He doesn't go on until 3.30. There's another speaker before him that I've forgotten, unfortunately. We're going to talk about it, X or something. Yes, yes. Zeon. Zeon, X is Zeon. How can you not love Larry? He's controversial. It's great for the bloggers and the media like us. But also, he's been around the block. He's been kicking those keynotes. He's, like I said, McNally's gone. Jobs is gone. Gates has retired. The old school guys are gone. He's the last standing. But the other thing I will tell you you've got to give him credit for is what he's done with the company. I mean, he took what was fundamentally a database franchise, did some apps on his own, and then went out there and did one of the more masterful sets of acquisitions. I mean, this company has littered, this valley is littered with people who bombed acquisitions, who've gotten them and had the people leave. Most of them failed. Yeah, exactly. And you look at what he's done with BEA, with PeopleSoft, you know. And Son. He's got the only software stack that makes IBM look over their shoulder, and that's saying something. You know, I had a conversation last night with a senior executive in this industry who was not an insider, but had just spoken to an insider, somebody inside Ellison circled, and she had said that, he had said that he's never been more excited about where Oracle is at. You know, that's interesting. I mean, the guy's, what, 66, 67 years old. He's outlasted them all. But to be excited when you're that age. And he's pumped, so hopefully we'll see that today. I'm sure we will at three thirds. Well, I think part of why he's excited, that was the part that did come through Sunday night. He's bought a hardware company, and it's very clear Larry's still got the hots for hardware. I mean, if you listen to that speech Sunday night, you know, all those speeds and feeds came through. Yeah, that's his boat. I mean, that's his race boat. Oh yeah, that's his race boat right now. I mean, that was a sun keynote. Yeah. Speeds and feeds. Look at HP. I got hardware, little subtle message there. I think. Oh, he went after the power stuff with the T4 too. He remembers IBM, he mentioned it over his shoulder, and I think that's the point. IBM is looking over their shoulder. What do you think they're thinking right now? For your IBM, you go on, again, they've had a transformation, got rid of their PC business, they're in the services business, but you know, computing-wise, they're there. What is IBM thinking from your perspective? Well, I mean, I think IBM is still going to pay out the power franchise. I mean, they've got a solid architecture, and you know, we were talking about it. It's a little bit like the foundation on which things are built. So you're not going to give up your foundation, but I would expect to see more software acquisitions. I mean, they've done a lot. They bought SPSS, they bought, I think the Nutiza move last year was a very bold move, given that, you know, prior to that point in time, DB2 would have been gold, and we wouldn't buy any franchise, you know, that was going to threaten that. The Sacred Cow, right? Exactly, whereas the Tiza, I mean, I knew the guys in Massachusetts who started, you know, some of the Nutiza founders, and it's a hot box. I mean, yes, they had lots of our chips in it, but I mean, really break the paradigm for easy use, load the database flat, none of these fancy schemas and indices. I mean, their promise was they'd go into a place and say, we'll give you a free for 90 days. And people would say, it's not even going to be running at 90 days. And they'd say, bet you, and they won, because you could turn it on and you could get that ease of use, which is, I think a little bit of what Oracle's trying to do with the appliance thing, which is give people a totally integrated solution that gets you up and running, because that's, people really want time to, you know, return. When am I actually going to have this box moving? Yeah, and that came out clear. I mean, Larry, Larry is obviously a salesman, he positioned the competition, talked speeds and fees, but I mean, these guys are a sales machine. So if you look at the keynote, I mean, my takeaway was, they just essentially froze the market, managed all the objections, one check box at a time, performance, analytics, in memory. So what's next for them? But also did it against the competition. You know, the T4 performance was against power, the exolytics was against Hannah. You know, I'm going to line my boxes up, bang, bang, bang, very specifically against the competition. I can't believe Larry's done with acquisitions either. I mean, you know, the sun one may have given him a little bit of heartburn and slowed him down a little bit, but he'll be back on the tarot, I'm sure. He's got a lot of cash out there. What's next for them? So given the check boxes, what's next? I mean, is there another race to run? Is there more check boxes for Larry to check off? Obviously the simplicity one, they got a lot of work to do mobile, not a lot of mobile discussion here. I haven't seen a lot of mobile. So is mobile it? Is it more expand, double down on analytics, bro that? What do you think is expecting for next year? Well, I certainly think double down on analytics. I mean, I think we've only seen the tip of the iceberg with the big data appliance and their discussion. You know, they did mention, was it Korean and that one that said, not only are we giving you the big data appliance, but we're giving you parallel loading because you got to put that data somewhere and you got to do something with it. So I would look to see more integration. You know, how does the data move across the various solutions they have? I think they're going to continue to be really competitive. I wouldn't be surprised to see them do more, you know, analytics acquisitions. I've obviously did the times 10 thing to get the in-memory stuff. He's going to look for places that he can fill out, techno, if I don't have it, I'm going to buy it because I'm not going to give anybody else a crack to get a foot in the door. What do you make of all those T4 supercluster claims? Well, do you know there are as many Zions in a supercluster as there are Spark chips because it's got the exadata storage and that whole storage subsystem is Zion processors. Yeah, well, we did a, at Wikibon, David Fleurer did a critical analysis of some of those claims. And I think you want to see the real facts. I think Oracle was a little economic with the truth on it. So the exadata is Zion and you got all the successes you guys had over the past. Just to go back five years, you guys had made a huge broad stroke across more markets. You got Apple chips and you're in, you mentioned Spark. Is there a hill that you guys have not climbed and dominated yet? Other areas that Intel needs that can go after that you think are still going to be battlegrounds that you need to fight? Well, I think we've made a very conscious, a couple of conscious decisions. We're currently a standards and a volume player. So, you know, we are looking for ways that we move the technology forward, but we're open. You know, we don't want to see cloud become a proprietary kind of thing, even though there are some purveyors out there that would like you to have their whole cloud stack. So we absolutely believe that the way the technology gets fully exploited is open. But the other thing we've learned, and it was a very, I actually wasn't part of Intel at the time. I watched it happen from the outside, is I think Intel learned a very solid lesson with AMD and Opturon. You know, it was kind of protecting titanium, not putting 64 bits, knew how to put 64 bits on Xeon. You know, it knew how to do that. And that's one of the reasons why when people say to me, are you looking at atom servers or you're looking at the low end, we're going to look at any way people want to exploit the technology because for us, it's how do you use it? I mean, does it matter to us whether people buy atom chips or whether they buy Xeon chips or whether they buy our many integrated core architecture, we want to put solutions out there that take Moore's law and deliver them for the usage cases where the high performance computing and the exascale guys are different than the database guys that are here are different than the client guys. So two questions, this is a look at how long, five minutes, damn, okay, this is so good. You can stay forever. So cloud, so let's talk about battlegrounds and things that are still up for grabs. You mentioned the people of different approaches. But first, before we talk about cloud and those battlegrounds, storage, which you know a lot about, cloud storage in particular, cloud storage has always been bits in here and there, but we're seeing companies like Nirvana, for example, using cloud at a petabyte scale for full motion video, you're seeing Warner Brothers in the studios pump up tons of video, obviously multimedia, we're pumping out live content that's being stored on blobs, a disk, moving across the cloud. What are the key things that are happening in cloud storage that still can advance in a Moore's law way to make it not so clumsy or make it faster or? Well, getting back to my SSD comment, I think you are going to see. Or cheaper, more inexpensive for us. Right, and I think that that is an area that's ripe for another generation of innovation and you're going to certainly see more of the solid state stuff. And the reason you're going to see that is because it's really, the volume space is driving the cost down. We wouldn't be plugging SSDs into servers if the cost hadn't come down because you all want instant on and the tablet guys are driving that. Thank you iPod. Yeah, exactly. But going back to what I said before, we haven't really started to exploit it yet. To just replace a hard drive is really kind of a dumb way to do it. I think you're going to see tiers of storage and you're even going to see some of it move in from the storage devices close. Why aren't there SSD cards that more directly plug into the server side as part of how you tier storage? Well, you're seeing that today. Yeah, and you're going to see more of it and you're going to see, I think you're going to see a lot of investment in the non-volatile technology and how you take it to a next generation. And that may take a few years to evolve but we're going down that space and we're going to evolve the storage side of the equation. And the interesting thing about non-volatile from my point of view is nobody talks about the availability aspects of it. If you could really have your whole first tier of memory in there being non-volatile, you wouldn't care about the machine going down because when it came back, it would give you everything you had. The whole issue of how you protect the transaction and don't want to get caught as you write it out to disk. We're going to write it out to disk anymore. You just write it to memory and it stays there. So it gives you a whole different thought process for how you think about high availability. The notion of state completely changes, doesn't it? Exactly, exactly. And is that a hardware problem, a software problem? I mean, I know it's both but is it largely a software problem? It's going to be some more of this database innovation. I mean, if we can give people much, much lower costs, much larger non-volatile, the software guys will have to think about how to use it differently. I mean, it's going to continue this rough. You know, when Larry says getting excited, I agree with him. There's more innovation in the database space. Database for a while was boring. You couldn't even get kids to take the courses, you know, for the computer science degrees. It was kind of like, you know, everybody kind of moved to Unix. Ingress, SQL, yeah, yeah, exactly. And now you're seeing just, I mean, look at Hadoop. Look at, you know, all of the memory stuff. You know, there's, and I think that trend is going to continue because I think the, as you say, this abundance of data is going to force people to look at new ways to organize it. And who's to say that there isn't some freshman at Stanford right now who's thinking about some new way to encode data and video, you know, so you can, I mean, the real thing people want for video is to analyze it. If you could actually run analytics on it, you could keep it a lot more efficiently. And those Stanford guys call us because we need, we want to hire you. My final question is we're getting the hook here. Pauline Nist at Intel, general manager of the data center group. Cloud, infrastructure as a service, platform as a service, software as a service. We all know software as a service is the big deal. Platform as a service is hot. It's all the rage. We've heard that at VMworld. Infrastructure as a service, a lot of converged networking stuff going on. What's going, where is the real deal? Where are the land grabs? Where are the battles still to be won in those three areas, in around cloud? And which ones are overbuilt or which ones are going to have the most consolidation? Can you just give us a quick layout of those three areas around kind of the state of the union there, in those three areas? Well, I think you're still seeing an enormous amount of explosion worldwide in public cloud, just because the kids today, I mean, they don't even want to talk to you in person. They want to text you from the other side of the wall, you know, that's what I love about it. People don't want to look at somebody's eyes and talk to them anymore. So you're still going to, you know, the amount of data, the photos, the putting your whole life on the internet, you know, I think the kids are going to still continue to drive that. When you get into our side of the business, I'll tell you one of the revolutions that I see changing that's already underway and you guys ought to go sniff around at it more with some of the ISVs that are here. I think that people who have been traditional ISV partners or what they've done is they've partnered with a hardware supplier and sold an installed solution to a customer are now looking at it and saying, maybe I should offer that as a software as a service. You know, why should I let a managed services company, you know, manage that configuration for somebody? If what they really want is my software and my solution, I know more about running that than anybody else. Let me put it out there as a cloud service. I'll make it redundant. I'll make it available. I'll give you the best performance. I'll drive for the lowest cost for transaction. And if you're in a vertical and you want to buy it, I'm going to go from being a traditional ISV to still being an ISV, but an ISV cloud provider where you buy my solution that way. And that's where I wonder, how does the traditional OEM ISV relationship evolve? Because that puts different people in the driver's seat. That's a disruption. The disruption is everywhere. The services model is changing. By the way, we have a new site called servicesangle.com head up by Alex Williams and we're going to track all these services just from deploying services, how to develop services, Pauline Nist, thank you for coming on to theCUBE. This is what it's all about. It's changing that insight and knowledge. Thank you for sharing. Thank you for coming on theCUBE. Appreciate it. I love to talk to you guys. I think it's great that you're live broadcasting. I think this is the way to go. I mean. We have 1400 people watching right now. Thank you guys for watching Intel. A lot of insight, a lot of speeds and feeds. This is the ESPN Sports Center of Tech, theCUBE. Okay, we got to get some web gems going. Like we got a new franchise there. Okay, Dave, Pauline, thank you very much. Thank you.