 Live from San Francisco, extracting the signal from the noise, it's theCUBE, covering Oracle OpenWorld 2015, brought to you by Oracle. Now your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live at Oracle OpenWorld for exclusive coverage of Silicon Angles theCUBE, which is covering Oracle OpenWorld 2015, hashtag OW15. This is theCUBE, Silicon Angles flagship program that goes out to the events and extracts a signal from the noise. Silicon Angle Media consists of SiliconAngle.com, SiliconAngle.tv, and wikibon.com research. I'm John Furrier, the founder of Silicon Angle. I'm John Michael, Dave Vellante, founder of wikibon.com. We're next guest, David Donatelli, executive vice president of Oracle Converged Infrastructure. Welcome back to theCUBE. It's great to be with you again. It's great to see you. We've interviewed you in many of your lives and your keynote this morning was really awesome. I want to say congratulations, your first keynote at Oracle. How did it feel? I mean, you got the new role, reporting to Mark Herd, you're up on stage, really kind of laying out the realities of today's market. So first comment on, what's it like to be the first keynote at Oracle? And two, the realities of the market today is a trend. It's not a one-off data point. You mentioned obviously the Dell EMC in context and you're looking at all the stuff at IBM and transitions at HP. Not a direct comment, but the industry trend is pretty significant. Yeah, so first of all, it's great to be at Oracle. One of the things that really excited me about joining Oracle is I thought that we were so well positioned for where the marketplace is going. And I know it's a cliche to talk about, oh, technology's moving very rapidly. But as it relates to infrastructure, which is what I've dedicated my career to, it is moving at a pace right now that is unheard of. If you sat here for this week and I know you have, you saw yesterday Thomas Curry and about four charts announced probably what was like 30 some products. Like I couldn't keep up. I was like, oh, there's another one. Here's a new one. Here's a new one. Off we go. And all that type of speed is having huge impact on infrastructure. We're in fact going through the biggest change, I think, in 25 years. And because of that change, that 25-year change, it's really impacting who's going to be successful as a provider, not only today, but going forward. And just like we saw the last time this change happened, which was when the mini computer died, the folks who were leading that back then and being a Massachusetts resident, I knew them all, deck, wing, prime, data general. They didn't make the transition to best of breed. Now we're transitioning out of best of breed to cloud-based and a lot of guys aren't going to be there as this thing progresses. I had a chance to sit down with you and Dave when we were at the cloud launch and then I wrote an article on Forbes in July, actually, months and months ago. And you really nailed it. You had a chance to go inside Oracle in your new role, look around, and just my observation, you were like a kid in the candies and you were like, damn, the products are good. But also you commented on this flattening of the infrastructure down to the chip level, which Larry announced. But more importantly, you said the comment, if someone doesn't have a public cloud, then they're in the box business. They're pretty much toast, that's my word. But I forget the exact quote, but what's happened between July and now? Obviously the Dell EMC things happened. You see a lot of transition in the marketplace. Customers want a new way to do things. What's happened in that time and you still feel the same way about this flattening? Yeah, well, I'd say two things. First, yes, I am a kidney candy store to be here. Because from an infrastructure perspective, I think what most of the market doesn't know is how much infrastructure development this company's done. And as you said, from the chip all the way up through the cloud, right, has been done in-house, good old-fashioned, ending in integration. And that's very exciting to be a part of. Second of all, as we discussed in July, I made a statement that more or less said, if you're an infrastructure developer and you're not successful in the public cloud or don't have your own public cloud, it's going to be very difficult to succeed going into the future. And there's a lot of reasons behind that. One, first is economic, right? If you're just an on-premise provider, as you all know, you read the paper, the economic model of making money that way is in decline. And as you also know, Wall Street's very impatient. They always want to see your economic model in ascent, not decline. So you're not going to be able to make enough money as a large-scale infrastructure provider on-premise. And then that leads to the problem. Once you don't make enough money, what do you do? You cut costs, you cut R&D, and therefore you're not going to be able to develop products people want in the future. However, in our case, we develop on-premise and in the cloud. That gives us the economic model that will flourish and it gives the opportunity to continue to invest in the R&D of the products people want going forward. You talked in your keynote about good old vertical integration. We're sort of back to those days at Oracle. But when you were at EMC, you were at HP, you obviously integrated with Oracle. You sold products into that market. It was a big market. What's specifically different about what Oracle is doing? Can you just peel the onion for us and compare it to sort of what the guys on the outside looking in do? Yeah, so here are some of the key differences. First of all, we have all the software stack here. We have the database, we have the middleware, we have the applications, and now we have the public cloud. So as a hardware developer, if you're developing the chip or if you're developing any of those elements, you can collaborate with the people who write the software to do really cool stuff that either increases performance, increases scale, or reduces cost. We gave a couple of examples at the show. One was a security in silicon where we even embedded things in silicon that enable your software to run more securely, either encryption, speed up, or memory that can only be viewed by one particular application, not others. Things like hybrid column or compression where we can compress at great rates, allows you to store less, also allows you to do things like query much faster, and when you start to use the two together, you get more benefit because in today's world, if you encrypt, you can now can't dedu, right? So now you need to compress. So these are just some examples of how working together is different. So those are two examples that your competitors can't do. That's right. So they're sort of stuck doing like RMAN integration and there's a level that they can integrate too. How do you decide where that line is drawn? Well, here's the basics. I mean, if you look at what everybody else is doing out there, it's I'm gonna take my favorite x86, my favorite storage, my favorite network, and I'm gonna put that together. And then on top of that, generally speaking, there's a Linux layer and then you have your virtual machine of choice. After that, you have no special insight into the software layer, you just don't. So what you try and do is say, well, I'll make this work as good as it can, then the customer can put the standard database or app on there and it's gonna run as best as it can. The difference between that and saying, well, gee, I have a private column to the database that only I have access to that speeds query or speeds the decryption. I call that one side of the bat phone to the database. But so, okay, so you talked about this pretty much the laid out a scenario for the end of the do it yourself best of breed in a world. However, does- I think that's better for customers. But so should we infer that can you, do you have to be best of breed? Can you be best of breed? Talk about best of breed versus integration. Is it a trade off? Well, you know, I think, well, here's what I'd say first of all, from a customer point of view, look at our engineered systems, right? I've literally talked to, as you can imagine, lots and lots of customers all over the world since I've been here. They all kind of go through a similar process. Hey, I thought your stuff was too expensive. We tried to build it ourselves. In many cases, reluctantly, we bought an engineered system. And what did they say? Ran faster, scaled higher, more responsive to IT, and on a total cost of ownership basis, saved a lot of money. And I've talked to customers, you know, I put up a slide today, talking about a customer who saved $42 million. I've talked to customers who saved money, bigger customers, obviously, and the hundreds of million. So it saves them money. It's actually the most efficient way to do infrastructure. And there's enough, you know, it's not theory. There's enough real-world cases to show it's true. I mean, that's a great example of customer banging their head on the wall, if I, okay, I give up, we'll go with it. And they try, they bring it in, so land and expand, which also, Sean Price was talking about, it's both on-prem and on the cloud as a growth strategy. So that's positive. But I got to ask you, with that thought of this new customer environment, whether they're banging their head on the wall, or they see clear shift that's upon them. You mentioned to me, when we talk, this architectural shift, because we asked, I asked you the bubble question, are we in a bubble? What's going to happen with consolidation? We talked about that. But the reality is, that trend is here, right? The management teams are the best, the best products that work in this new environment will win. So I got to ask you again, what is that architectural change fundamentally that is creating this new era of different looking products, different integrated stacks, you're looking at more agile, more DevOps, infrastructure as code, yet security plugged in there? Can you comment, because the engine of innovation, I mean, it's like Oracle's like a Tesla now. Yeah, there's cars out there, but this is a new kind of engine, this shift, talk about that architectural shift. Within ourselves or within the industry? Oracle and the industry itself, the customer is what they're facing and how Oracle ties into that shift. You know, you heard Larry say the other day, which I thought he put a well is, the best way to understand the shift is look backwards a little bit, right? And look at the pieces that have been assembled here over the last 10 years. They've redone all the software. They either built their own infrastructure or acquired infrastructure. And then they built that into a completely comprehensive cloud, infrastructure as a service, up through PASS and SAS. All put together over a long period of time, whole new infrastructure. And regardless of how you got there, that's the way the market wants to consume. And that consumption is driving the change in the way architectures are developed. As you guys know, for 25 years, what was it? I'm gonna take apart each piece and build it myself and on my particular company, I'm gonna have this knob different and that knob different. What's cloud? The complete opposite of that. I'm gonna take it as it is and I'm gonna use it more efficiently the way it is because I can get up and running faster and it's more efficient for me. So this is not an incremental improvement. The old days, okay, add a new feature and now it's a game changer. All right, now let's go back to the architectural question. Let's go a little bit deeper under the hood because it was great. Now all the inner geeks of the hardware guys were going, yeah, love the engineers, this baller. In memory is a big here, but flash has hit the scene hard. You see pure storage out there. So flash has changed the game. So talk about what's going on with the architecture around flash and now within memory. How do customers make sense of that? What should they be looking at that from an overall deployment standpoint of solutions? Yeah, well, certainly as you guys know, right? If you look at flash, that's we have everything changed because of cloud. That's what's charging the big architectural change. Then if you kind of go down to the storage layer, they got other changes on top of that, right? Flash is one of them. As you know, that market, the analysts say it's two billion this year. It's gonna be four billion next year. It's growing like a rocket. And then that's forced everybody to redesign from the ground up to take advantage. Traditional rays, as we know, yeah, you can put flash drives in them, but they don't have enough IOPS within the bandwidth of the array for customers to see it. So what's that force? To your point, ground up flash arrays. We built one here as well. It's called the FS1. The difference between ours and everybody else's is we also tied it to our database or I've been aware in our cloud. So you get that full stack integration on the way through. It has like a broken record on that, but that's the strategy. I mean, that's what it is, right? But it's a pendulum, David, because, you know, if you look at it, why did people like mini computers 25 years ago? Because they wanted a full stack. That ran its course. We then spent 25 years ripping the stack apart. And what we've done now is we swung the pendulum back. And I think a big driver of adoption is actually people's personal lives. You know, you have your phone, you have your app-centric world. You're used to doing it personally. And you say, I want the same thing in my business. They're shrinking the stack too. They're actually shrinking some core stack and adding new stuff onto it. This new integrated piece. Yeah. Well, it's a fundamental part of your strategy, obviously, that hardware and software integration, but it's also a component of this. Well, it's very different than what's been going on. Yes, it's exact opposite. It was because it was 180 and it's working. But so, but you also draw a line between what you will let competitors do and what you guys will exploit. You've mentioned a couple. Hybrid-column compression, OISP, the stuff you're doing in ZDLRA. Those are all unique. Yeah, but I want to be clear about this, right? For years, our software has been open to all, right? Our biggest competitors in hardware have been long-term Oracle partners on our database, on our middleware, on our apps, and they always will be. And they can still run on Oracle. Absolutely, right? And the only difference we're saying is where we invest our specific IP dollars in developing our hardware. Yeah, we're going to enjoy the benefit of that. There is a but there, Dave, right? They can't do that. We're going to enjoy the benefit of our development. Sure, but they can't do that. We spend five and a half billion dollars on R&D for a reason and we should get benefit from that. Well, so, but it's hard to argue that stuff doesn't run better on Oracle. I agree with you. Right, okay, so it's a long-term strategy to, I mean, that's increasing. The number of examples you can point to where your competitors can't is increasing at a pretty rapid rate. That's right. And we should expect that it's going to continue. Well, here's why, though. What I contend, though, is because we're investing so heavily, right? It's not that others don't have an opportunity, although unless you have your own software stack, it's difficult. They can't, right? They can spend all the money in the world. But you have to have the software stack and then you got to be willing to invest like we are. And we're investing extremely heavily, not only just on our software and cloud, but what we're doing in infrastructure. I think people will be shocked. I always, I welcome anybody, you know, who's interested in infrastructure to come see us and in particular go to our Santa Clara campus because that's where we develop all the infrastructure products. And if you're into that kind of thing, if you're into chips and networking protocols and storage, I guarantee you'll be blown away. How many engineers, what they're working on and the depth level they're doing. And I think when you saw the M7 announced this week, that's an example of that, right? That is not an inexpensive development effort. Right. And it's just symbolic of all the other things we're doing. Well, I mean, I'm on record, John, you know this for years. I said Sun was one of the best acquisitions in the history of the computer industry. I got criticized for that. And it's playing out exactly as Larry sort of laid it out. So I got to ask you, what's a better name? The zero data loss recovery appliance or the Oracle private cloud machine for infrastructure as a service and a platform as a service? I go with the second one, but I won't repeat it due to time limitations. And you know, and they joked about the keynote that Larry named those. And I could tell you how I was there for one of those, not the ZDLR I was named for, I got here. But it's absolutely true. He names them. And it's awesome being at a founder-like company because a founder-like company, one, you get the founder's vision and technology, but you also get great clarity. Yeah. Because when he says that's the name, everybody says that is the name. So of course I'm joking, but the ZDLRA is a game changer in our view. We just did some research on this in terms of backup and recovery. Yes. I wonder if we could talk, I mean, you're doing stuff in memory, your end-to-end stack, again, the same theme. Like I said, it sounds like a broken record, but it's changing the way in which backup is done, backup windows disappear. It's essentially a time machine for the enterprise I've called it. Yes. Talk about that a little bit. Nice term, by the way. But you know, here's my view of the whole thing. I started working, one of the first big software programs I did was backup software, and this is about 20 years ago. And I'm either humbled or saddened to say that in 20 years, backup has still been very difficult. The big leap that's happened was moving to backup the desk. That at least got it faster for people and able to make their windows. But I can tell you, I've had to do customer support. You're on the phone late at night, there's problems, and it is shocking. I've been there, I've been on the phone. People, like, I don't have a recovery option for that. You know, and I still talk to customers all the time who run into issues, how do they recover their most mission-critical stuff? The reason why I think ZDLRA is so important is, you know, it was written by the guys who wrote the database. And if you think about that, as for as simple as that said, it's like, why wasn't that done 20 years ago? Right, the guys who wrote the software actually figured out how to back it up and recover it and recover it to any point you want. So by having the guys who actually wrote the software, write the, you know, it's an engineered system, but it's a lot of software, write the software to figure out how to recover it at any point, probably the best chance of having it work great. And that's why I think it's so strategic. Now it's got all kinds of great benefits of automating backups, automating restores, making it, you know, trouble-proof. And I think it's a game changer for a ton of customers on what is their most mission-critical data. One of the things we talk about all week is that Oracle's positioned across the board some good territory, they're right in the middle of all the action, but they've got two vectors of competition, if you will, or customer concerns. From the top of the stack down SaaS applications and bottom up with infrastructure as a service, AKA Amazon Salesforce Workday on both sides. So what are you, tell me more about the cloud services, some of the things that you're doing in infrastructure as a service, because there is an engine of innovation that's coming out, the M7s, we talked to John Valer, encryption, all that great stuff, but what is the concern for people who are out there looking at doing their own infrastructure service or going to Amazon? What are you guys doing? How are you differentiating on that front? Yeah, I think if you look at, you know, versus like an Amazon or any of those other folks, the biggest thing, there's a lot of things we do, but let's just talk some of the basics. First, we offer engineered systems in our cloud. As you know, all these other clouds are white box. So for people, you know, we're obviously believers in engineered systems, you can run your engineered system on site, you can run it in the cloud, you can run it cloud only, having that high end capability in the cloud, very unique to Oracle, number one. Number two, being able to move seamlessly amongst all these, meaning from your premise to the cloud, very unique to Oracle. And that gets back to business model, right? There's three cloud business models out there. There's people say, I only do private cloud, I don't do public, you know, we'll help you go to somebody else. There's people say, I only do public cloud, I don't do private, so good luck on your own premise, you got to figure that one out. We do both. And since we do both, we made it very easy to traverse both worlds. And I think, you know, although there's a whole bunch of customers out there with all different needs, virtually anyone who's been around for some period of time is going to have needs for both, not one or the other exclusively. And the things we're hearing about from the customer feedback is that what they don't want is they don't want to be in the integration business. That's right. They don't want to be managing the stack because it's inefficient. Can you comment on that piece? Well, what I believe is that those other models I refer to keep customers in that integration business, right, because you're traversing, you know, I said- So you think they should be in that business? No, I don't want them to be in that business. But I think what makes our model unique is we can help them get out of it. We have the simplest to deploy cloud. If you saw Larry's keynote yesterday, he moved the database, you know, from Prem. Right click. Yeah, right click, done. Okay, simple. Now, I challenge you, go do that somewhere else and see how that works. Everybody likes to talk about cloud bursting, stipulate. Everybody's cloud bursting probably technically works. Now, go try and do it practically. The other really important thing you just said is that Oracle runs optimized Oracle, Oracle optimized hardware in the cloud. You're the only ones who do that. That's right. So unless Amazon starts buying exodatas That's right. and configuring them, are they doing that? They're not. Right, and then so what does Amazon run? As you know, they run all white box, white box server, network, storage. How many enterprises run pure white box, server, network, storage? Can't think of many. So by definition, you're running what you're running on your enterprise. You're running, Amazon's running something completely different. Then in that scenario, you as a customer have to figure out how to navigate two completely different worlds. In our scenario, we make it simple. Same world, both places, same software, both places, same skills to run in both places. Okay, thanks for coming on theCUBE and sharing your insight. Dave, do you have any final questions to get into the last couple seconds? Yeah, so, last thoughts on being here. You know, when you keynote, you talked about your history and where you've come from and now you've gone for the full circle to back to vertical integration in the 80s, you know? It really is full circle, right? It's where my career started was vertical integration, although I was the one who was against that. Yeah, yeah. You were strong saying you wanted to be best of breed. But no, it's fantastic to be at Oracle. It's a great company, great culture, and it's a really R&D-centric company. It's about building great products that help customers with their biggest problems. That's a lot of fun. My final question for you, Dave, is cloud consolidation. Will there be three clouds? Is it winter take all, winter take most? What's the number that finishes that, you know, as we settle the dust a few years down the road? Depends how you want to count them. But if you look at major, you know, big enterprise kind of things, they can count on one hand. One hand on the cloud. That's Dave Donatelli's prediction. Obviously, Mark Hurts said 100% of tests and that will be in 2025 in the cloud. Oracle has got the cloud business. Congratulations, congratulations to the new role. This is theCUBE live on Howard Street, where they're shutting everything down for 60,000 attendees. This is theCUBE live from Oracle Open World. We'll be right back with more after this short break. If you want to see all the content from the videos, go to our new crowdpages.co slash oow15. We have all the videos up there in real time and all the conversations from CrowdChat and all the trending stories. We'll be right back with more after this short break.