 We're back, this is Dave Vellante, and we're here live. This is day one of Oracle Open World. I am sitting in for John Furrier, who was driving up the Pacific Coast Highway, dropping off his son at the University of Santa Barbara, and he'll be here this afternoon. I'm here with Stu Miniman and Jeff Kelly. Gentlemen, we've been hearing a lot about Oracle. We heard the keynotes this morning, Larry Ellison last night announcing an in-memory database. One of the things Jeff Kelly, we didn't talk about in depth was that in-memory database that was announced, it was really a response to SAP HANA, wasn't it? And some of the other trends that we're seeing in the industry. What's your take? Absolutely. So essentially it's an in-memory database, a little short on details, but an in-memory database that allows you to do some analytics and transaction processing on the same database, which is exactly what SAP HANA is really designed for. So I think certainly Oracle, they continue to respond to market pressures, SAP HANA being one of them. But of course also with a larger big data story responding to the pressures associated with the open-source big data movement around Hadoop and other NoSQL players, which we're going to have on later today, folks like DataStacks and we'll talk a little bit about how they're competing with Oracle and really trying to add a new element to the database game. So there's an article in SiliconANGLE that I just saw, I think it was Bert Lattimore put about talking about how Clustrix and I think even IBM had come out and said that Oracle's approach to in-memory is not the correct one, that Oracle's doing a scale-up, that the future is scale-out. Stu, this is a big argument, the whole scale-up versus scale-out. We have this all the time. Clearly if you're Amazon, if you're Google, you're going to move into a scale-out. You've got the PhDs and the engineers to actually build a software-led layer that you can put on top of commodity infrastructure. But Stu, most enterprises don't have that capability. Most enterprises would rather buy a solution. They'd rather spend money on a solution to save time, whereas a lot of the hyperscale guys, they'll spend time with engineering time to save money. So what's your take on this whole scale-out versus scale-up thing? Yeah, David, exactly as you were saying, not every customer necessarily needs to just continually scale. Absolutely if you look at the service providers of the world or the large customers that act like a service provider, they are growing and they have a growth profile where they need to keep adding and expanding. And what we don't want to have to do is every time I outgrow one box or one configuration, have to add that into my network and continually expanding and making it more difficult. We do want kind of seamless, fast growing, which is what scale-out architectures really allow. For many customers, though, a single box that they can grow inside is going to last them that kind of three to five years on a refresh cycle. So the real thing we look at is some customers are going to be OK just buying that single box solution and upgrade the pieces of their stack as they go. Others might need the scale-out architecture and everybody is struggling with what do I do in-house versus what do I put in the cloud? Because that's really one of those discussion points is, is IT a differentiator for my environment or should I just be getting it out of the purview of my own people so that I can trust somebody else either for a specific application or for smaller companies that might even be my whole IT environment? So a lot of times we're talking about in the cube about Larry Ellison has Apple Envy wants to be the iPhone of the enterprise. But so what are the similarities and what are the differences between what Apple is doing in software and engineered hardware, I guess, as well, but particularly Apple as a software company and Oracle? Jeff Kelly. Well, I think if you look at, well, it's interesting if you look at the iPhone, for example, obviously it's they've had a big refresh of the software iOS 7, but also released some new hardware. But if you think about the things that make the iPhone so attractive to a lot of consumers as it's simplicity, it is, you know, it's elegance, but there are some challenges associated with that. You know, for instance, it's expensive. You know, you're paying for all that preintegration that took place. There's also some, you know, it's not as customizable as an Android device. I mean, we all everybody's got an iPhone is familiar with, you know, the new stand app and the the stocks app and others that you don't want to use, but you can't get off the iPhone. And similarly with Oracle, you've got to accept certain realities when you adopt their their technologies, their converged systems. So in that sense, you know, they are pretty similar. Of course, Larry Ellison and Steve Jobs were, you know, close friends. I think saw eye to eye to a degree on this kind of idea that hardware and software should be preintegrated and really built together. Is the let's let's stay on this theme of mobile and and consumerization of I.T. You look at the iPhone totally integrated Android is this Android is almost like the windows of mobile devices. And now Microsoft has gone out and it's bought a device maker and it's gotten into the to the hardware business. Google has done the same thing. Stu, is the analogy of the mobile marketplace and the trends that are going on in the market. Are they applicable to the enterprise or are they just two totally different worlds? Interesting question, Dave. As a matter of fact, you know, Scott Low, who who writes for us at Wikibon, just wrote a piece that just went out that, you know, maybe Microsoft, you know, should look at buying Blackberry. They just bought Nokia. And, you know, when Microsoft first got into the service, I was really skeptical about Microsoft getting into the hardware business because so much of their business was driven through the OEMs. And I thought they really were putting surface in place to try to incent people to adopt, you know, the Microsoft platform and that they would get HP and Dell and others to kind of drive that platform. So on the analogy I draw on the enterprise side is when VMware launched their vCloud hybrid service, it really, you know, had a lot of pushback from some of the service providers because VMware is a technology supplier and has a huge ecosystem that either sells the licenses that VMware offers or builds solutions around them. So it is really tough for a company to span between them. The great example of who has done a good job of being able to kind of eat their ecosystem without losing all of their partners is Intel. And Intel is one of the few that's been able to do that. You can buy Intel white boxes, but they still compete with them. We had Sandisk on earlier, who makes it very clear that they are a supplier to many of the providers out there. But it's tough to be both a supplier and a competitor in that space. And, you know, we've seen kind of mixed results depending on how you deploy it. So we've heard a lot about cloud. We've heard been hearing cloud from Oracle now for quite some time. All this talk about the NSA and and Prism has really heightened awareness of security and privacy. Doesn't that bode well, Stu Miniman, for Oracle? Yeah, David, it does. Because if we look at, you know, this is not a story that's going away anytime soon. We've talked to a lot of CIOs and security is all of a sudden, you know, one of the top priorities and people are worried that if they go to a Google or a Microsoft or an Amazon, that they might be compromised. Heck, all of the encryption technologies that people are using in house, they're all worried about whether there's a back door that the government can see what's doing there. So, you know, this all means that I'm more likely to keep things in house and keep things under my control. And of course, that is where Oracle has a good play and the traditional infrastructure players, you know, are less likely to be disrupted from the likes of Amazon. So do you think we're going to see Oracle, you know, putting some flood into the market about this security question? That's a really interesting perspective. I mean, if they've got a lot to gain from this, Oracle is not above, you know, stirring the waters a little bit and putting a little foot out there. So definitely anybody that offered private cloud solutions has been saying for years that you can't trust the public clouds. And in many ways, we thought over the last, you know, 12 months that people had gotten beyond that. Amazon has their government cloud and there were solutions that are going out there. Everybody's been testing in the cloud environments. I mean, Jeff, I know you watch all the big data environments that go out in the cloud, but there's definitely foot out there, but there's reality that people are worried about, you know, what is the government going to see and where do I put things? Well, I agree. It'll be interesting to see over the long term if this, you know, whole NSA scandal, if you will, it really stifles innovation because the cloud offers so many things you can do, particularly, you know, in the area that I cover big data. It really, the cloud is going to be critical when it comes to things like the industrial Internet, connecting the internal infrastructure of the country, the electrical grid and other things with healthcare and other areas of that nature. And so if that kind of stifles investment in or innovation around the use of cloud to enable industrial Internet use cases, that would be unfortunate. So I want to go back to the big data themes that we heard this morning. Oracle was really beating that drum pretty hard. I mean, Jeff, to me, it's a total, you know, transformation of the whole big data theme. I mean, last year we saw Oracle. Larry Ellison gave this big demo. He showed, you know, it took the Twitter fire hose and he ingested the whole thing into, you know, exodata or exolytics. I can't remember which big box it was, but it felt kind of million dollar iron-ish and Thomas Curian today was more talking about a portfolio, which again, I think would resonate well to a CEO or a senior level CIO who's hearing all this stuff about Hadoop and big data and unstructured data and data growth and data being the new source of competitive advantage. And then Oracle comes in and says, look, we got this portfolio, it fits with your existing legacy environment. We can wrap a nice red blanket around you. Here you go. And you can solve all the world's analytics problems. Why doesn't that play well? What's the risk that customers face in going down that path? Well, I think the major risk, of course, is that you get locked into an Oracle environment. And if innovation in the open source world kind of continues at the pace we're seeing, potentially you get left behind. I think, you know, obviously the other thing is it's going to cost you significantly more to invest in the Oracle's big data products, solutions, architecture, what have you, than going out and spinning up a cluster on AWS for Hadoop or bringing in, buying a few machines internally and spinning up a cluster, certainly it's going to cost you a lot more. But again, Oracle is betting on that, hey, look, the fear of lock-in that hasn't stopped us from growing, hasn't stopped our customers from investing in us, and they're betting that it's not going to stifle their big data play. So here's the conundrum, 50% of the customers that we surveyed in the Wikibon community said they were willing to risk lock-in to get integrated function and simplicity. That was sort of a clear message, and only about 15% said that we are dogmatic about open source. So open source in and of itself, and this is not a surprise to anybody, clearly is not the driving factor for customers. What is the driving factor is innovation occurs as a result of open source, and what you're saying, if I understand it correctly, Jeff, is over time the open source community will catch up and actually will surpass sort of the proprietary vertically integrated stack and then extend value at a much lower cost. Is that correct? Absolutely. Look at some of the work that, for instance, Horton works is doing on the Hadoop area, with Yarn and MapReduce2, you know, they are really pushing innovation in that area, and they're also offering a value proposition that's hard to beat, the software is for free, you pay only for support. They're starting to integrate with other providers of data warehouse, data-based technology like Microsoft and Teradata, so they're trying to make it easier to integrate into your environment. But really, if you get locked into an Oracle, big data play, and you could potentially see the open source world and other competitors zoom by. That said, Oracle is not, certainly they're not stupid, they're planning, they have a good strategy around this, and clearly there's going to be some acquisitions in this space in Oracle, I think is going to be one of those companies that make some of those acquisitions. So, you know, it's really, I think they're their play around integrated systems that allow enterprises to start driving value from big data today, I think is going to play pretty well with their customer base, and it's going to certainly play well with a lot of folks out there who are struggling right now to bring together all these open source technologies and haven't had a lot of success yet. Okay, well so same question on the hardware side, Stu Miniman, how long does it take for software-led infrastructure on commodity components to actually catch up to the V-Maxes of the world? Yeah, great question Dave, it definitely, you know, these kinds of transitions take a long time, and as you know, we talked about earlier, customer environments in the enterprise are really looking for full solutions, so it really does take somebody to bundle all of those pieces together, because most customers aren't going to be able to take, you know, just, you know, take software and buy their own hardware and wrap it all together, which is what we see Google do, you know, Google puts their entire configuration together, they did their own distribution of Linux and put it all together, they've got, you know, all the PhDs to make that happen, there are interesting technologies out there, both the storage space and in the networking space, trying to disrupt them, but we usually measure those transitions in years or a decade or more to be able to do this, so, you know, it's interesting to see on the, you know, the storage side, we've been watching close, you know, EMC has their projects like Viper and Nile and what they're doing, but you know, it's going to take a long time, it's basically what I'm saying Dave. All right, Stu and Jeff, thanks for helping me out with this editorial segment. We're going to stay on this theme of Flash, Fusion IO is here, interestingly enough, when EMC sort of landed its haymaker and announced that it was bringing Flash back into the storage array, at the same time Fusion IO was working on a PCIE card and essentially shipped that product, you know, right around that same time, Fusion IO is the leader in that market, Gary Orange-Teen is here, so keep it right there, we're right back, we're going to unpack what's going on with Flash at Oracle OpenWorld, this is Dave Vellante, this is theCUBE.