 This is Dave Vellante, we're live, SiliconAngle.tv's continuous coverage of VMworld 2012. We're here with a couple of CUBE alums, good friend of the CUBE, Rich Napolitano, who's the president of EMC's unified division, and Eric Herzog, who's the senior vice president of marketing for that division. Gentlemen, welcome to the CUBE. Good afternoon. Good to see you guys. Thank you for being here again. Things are good, you know? I mean, Rich, they keep coming at ya. Keep knocking them off. That's right, that's right. It's like Andre the Giant with all these guys coming at him. And you guys are still king of the hill. Congratulations, by the way. Wikibon just released the results last week of its VMware integration study. Independent study of the quality of VMware integration. You guys came out number one again. We're very excited about that. The independent survey of customers. We asked, you know, who's the number one storage supplier in VMware? You guys, we asked customers that were not EMC customers who was number one, and EMC came out number one in that segment. So, across the board, you know, congratulations. No, no, we're very excited. And you guys do a great job on that survey and it's the voice of the customer, right? So, there's no greater truth. Yeah, right. There's no greater truth. Yeah, we're proud of that. We put a lot of time into it. Like you say, it is what it is. It doesn't lie. Yeah, no. And we invest a lot in that space and we'll continue to do that. There's some innovations that are coming this year and next in that space. We're going to continue our leadership. It was the end of the day. If we created, you know, tight solution with VMware, we would use complexity in the data center. And that's one of the key propositions for the market. Reduce complexity, reduce cost. And that's fundamental to really improving IT as we roll forward. So, complexity has been this kind of perpetual inhibitor to growth in the industry. Do you think we've solved that problem? Well, we'll never solve it completely, but we'll make it better. And I think, you know, with every innovation comes the introduction of some complexity. So, the trick is leveraging innovation while taking complexity out. And there's some things that we announced in our DMC world, which I think we'll talk about perhaps a little later, you know, vCenter ops, and just how do we make our technologies, whether it be flash or virtualization or management, how do we remove complexity to get people to receive the benefits of the underlying technology? That's the trick. So, get the benefits of the technology in terms of cost, cost for transactions, cost for capacity, cost to deploy a VM, you know, cost to deploy a seat, take cost out, but make it easy to consume. So, Eric, what's your point of view on all this? I mean, you're relatively new to EMC, a couple years now, right? Yeah. And Ccommon is an outsider. There was actually a lot of change going on in the unified division when you joined. So, from a marketing standpoint, what have you done and where do you want to take it? Well, our key thing at Unified is to drive application integration and application centricity. Everyone can talk about IOPS and gigabytes a second, but not many people can say, here's why our product is best for VMware, here's why our product is best for SQL or Oracle or SharePoint or wherever the app is, including vertical market apps. You know, 50 some percent of all healthcare institutions use EMC storage, for example, integration with a total different set of applications that most horizontal companies wouldn't be aware of. So, our key is focused on application integration and still continue to meet the needs of the storage admins and the storage architects, but at the same time, meet the needs of the CIO and the application owners with the same set of tools that allow them to easily integrate, manage, and monitor and optimize the storage that we provide. That's a critical thing for unified storage. So, Pat was on this morning, I know Rich, you just got here. You're old boss, I guess, was that there? I'm an old boss. Well, you never lose Pat as a boss, that's how it works. There were guys that didn't tell it still worked with Pat, actually, I should point out. And that's through EMC proper now in VMware, they still work with Pat. You're a legend, there's no doubt about it. But he talked about this museum of legacy systems and, of course, it's VMware's objective to bring all those together. Are you seeing organizationally those changes take place? I mean, Eric, you were just talking about all these different sort of application areas that you guys go after. Is that changing or is it just becoming sort of one ubiquitous IO blender or do you see that sort of staying fairly heterogeneous for a while? I'm not sure I understand the question. So, let me take a shot and then, okay, I guess we could. There's no question that virtualization provides an environment where you can collapse many applications into a common infrastructure. So, we certainly see that where many different types of apps come onto a common infrastructure. I think you had the New York Stock Exchange on here a minute ago. And they talked about very much the same thing from a kind of service provider to their clients. And they have a variety of different applications they build onto a converged infrastructure. The opposite of that is that some applications tremendous scale have discrete infrastructure. So, if you look at certain large scale block applications, they'll only use block infrastructure like a VMAX. If you look at certain large scale genomics, et cetera, which are file oriented, they'd use Isilon for very, very large scale file implementations. When you look at kind of unified, we live in this kind of hybridized world of kind of block and file, transaction and capacity. And so, many things that have a diverse set of applications on a common infrastructure would be on a kind of unified device. Is that what you were going at? Yeah, okay. So, now how does Flash affect all this? So, Flash is important because Flash is one of those innovations that, you know, we're probably 10, 15% into absorbing those technologies into our products at this time. And so, when you look at Flash, one of our kind of key adages is a little bit of Flash can go a very long way. What we mean by that is you can buy a little bit of Flash and have a dramatic implications on your application's performance. So, some 5% Flash, you can drive 80% of your IOs to that small percent of the Flash, but more importantly, drop the average response time for your applications in half. In other words, reduce response time to your apps literally by 50%. And that has big implications on people's applications. And you couldn't do that with just spinning this. You just can't. Yeah, it has implications on things like VM density, so you're going to attack the operational efficiency piece of it, but it's also going to start creeping in, isn't it, to the business productivity side, certain things you can't do with spinning disk. Correct. That's exciting. It's kind of scary. No, it's beautiful. It's a beautiful thing, right? It's a good thing, right? If you own the assets to be able to do it. And we own all the assets, so we have VMAX, we have Icelon, we have Project X or Extreme IO, we have our hybrid arrays, you know, VNX, Icelon and VMAX, and we have Lightning, we have Thunder, you know, VF Cash. So at the end of the day, we have all the host footprint of, say, like a Fusion IO, we have all the external stuff of IELIN, we have all of the kind of all flash arrays of Pure and some of the startups in the space, and we have a huge portfolio of these hybrid arrays. So at the end of the day, we have every single technology base covered, and which really puts us in a unique position in the market to actually be able to position not a product, but the right solution, because we have everything in the arsenal. There's no one else in that position in the marketplace. Yeah, so how do you see that solution evolving? Does it become more hybridized, more consolidated, more, you know, it's a big thing about converged, or do you see still very specific requirements for different use cases? Well, I think, you know, we talked a minute ago about complexity, and part of the challenge of having this rich portfolio is the complexity that it represents. So one of the things that we work on a lot in the unified storage divisions, and now even across with symmetrics, is to get into things like common management and common replication. So, say, symmetrics and the mid-range now have unispheres, common management. We also have recovery points on common replication. So we drive commonality of those touch points that are really, really important to the customer. Our APIs are getting more and more consistent about how we plug into things, like vCenter operations manager, et cetera. So underneath it, kind of right tool for the right job is our strategy, and understanding the use cases and the technology, but we need to simplify how we articulate that because it's going to be a little overwhelming. So, Eric, I wonder what you'd make of this? I don't know if you were in the keynotes this morning, you may have just arrived yourself, but there's a big discussion about the Software-Defined Data Center, which is all about abstracting the physical, logic from the physical, pooling the resources, and then automating. And there's also been some discussion about software-defined storage. So what do you make of all that from a marketing standpoint, messaging? Is it just sort of industry buzz? Is it real? Well, we see that actually with the unified products is fitting right into that. So VMware has already talked about VVol, and EMC will be their day one with VVol support, which is sort of their first progression. There's always going to be hardware underlying that software infrastructure, and while some people think that will commoditize, what has been proven that if it was all commoditized, it would be naked flash. If it was all commoditized, it would be basic naked disk drives. And quite honestly, people haven't seen that for 20 years, yet EMC and its big competitors are still all in business. And it's sort of proven that when you do these levels of abstraction, you still need tight coupling of what's underneath up to that level of abstraction. So Software-Defined Data Center, as well as Software-Defined Storage, and all of those, will still need a strong storage underpinning from a physical perspective. And what's key is the right stitching between the hardware underneath and how it operates with that software abstraction. So we actually think it's an ideal opportunity, actually for us. And we sort of proven that, for example, with our VMware integration already today. And with integration with other technologies and other software vendors that we share, is we can, as they progress forward with this, we still think we will be the strongest underlying physical devices that will have to stay there. And we will optimize our software and our RAID stack to optimize whatever they've got on the software layer. And bottom line, using a bare naked flash drive or a bare naked disk drive probably won't cut it. It hasn't cut it for 20 years. And the fact that EMC and all of its primary competitors are all still in business and doing well, sort of shows that it really won't change much. It'll just get faster and faster and faster. So let me reinforce some of what we just talked about. You know, these, as Eric talked about, these trends have been going on for a long time, but more importantly, if you take it to a big scale, you're still gonna need core networking. You're still gonna need core networking, switching and storage, right? Because the expectations for centralization, availability, failover, et cetera, big pipes, big infrastructure, you're gonna need that. So if you look at these things, they have to work together. And certainly you have to not, to enormous install base of all this infrastructure. And so very often when we start to find these standards, et cetera, the world wants to go from black to white or white to black and just kind of draw a strong contrast. At the end of the day, these markets are growing so rapidly, there's more than enough space for these technologies to coexist and coexist and must, right? So when we think about software defined networking and storage, that's gonna take a piece of the market and Eric talked about it. We've always had direct attach out there. And so those use cases may expand, but the fact is those, these markets are growing so rapidly that they'll need to be synergistic. I think what's most important about software to find networking and storage is around the control path and the management because that's one of our biggest challenges. So how do you integrate that? How do you make it seamless? Eric talked about and you talked about the awards around VMware integration. That is fundamental to taking the complexity of the infrastructure. So at scale, there is no question you will have dedicated networking and storage infrastructure. And then you can talk about the great agents in between and how you do that and then how do you deal with an install base. But I think there's more than enough opportunity but innovation in software to find networking and storage is fundamental to attacking the complexity of the infrastructure. And it's a system. I mean, you're right, Eric. I've been here for 20 years. Ah, hardware is irrelevant. It's all commodity. It just keeps getting more and more important. It keeps growing because it is a system. Well, I think what happens is when people sort of prognosticate the future, what they prognosticate is we need it to be simpler but at the same time it always ends up getting more complex. So when you look today for example at our hybrid arrays, we're addressing two fundamental problems with the same solution, which is the need for incredible IOP performance. At the same time, the need for IT, whether it's in the mid-market, the enterprise, even in SMB to drive the cost per gigabyte down at the same time. And those are fundamentally diametrically opposed goals. Yet with our hybrid array, the VNX and the coming VNXE, we can actually solve that problem and deliver the IOPs they want which is going through the roof as everything becomes more and more virtualized and more and more applications run and take big data, everything at the same time, we're driving the cost down at the same time. So people who forecast it, software's going to overlay all the hardware and commoditize everything seem to forget that as more and more apps become incredibly complex that you need that power of that hardware coupled with the right coupling, interstitching between the virtualization layer to come together. And I don't think that's really going to change. If anything, more and more and more and more apps are going to get more and more and more complex. So it's going to need pretty soon, two million IOPs won't be fast enough, right? And people will say, oh, you only got two petabytes installed, would it? Got a small config, people laugh, but I was back in the day, five megabytes was a big disk drive. So I remember that too very well actually. Now you guys are in the heart of this disruption and I want to get your point of view on something else which is for the last 15 years we've seen function move out of the host, serve into the array and that's been, you guys have capitalized on that and delivered value through a variety of different functions whether it's replication or snapshotting or BCVs or whatever it is. And you're starting to see with flash function move back. And Rich, you talked about the control path and the management. So today, in the last 10 years or whatever, IOs been a scarce resource and you really had to manage that quality of service with flash and all flash arrays and the like. IOs maybe are no longer the scarce resource. Are we going to be managing those from the server view, from the array view, yes? I think it's yes. But don't you need a single point of control? So you need to really take the natural extension of the things that I think we've been doing which is you really look at the manager in the mid-range, Unisphere and VNX, we really approach it two ways. From the top down with our deep integrations with VMware and then from the bottom up which is kind of the element managers up and crossing into the virtualized world. Because at the end of the day, the storage subsystems are special purpose built devices and they have unique properties. And so the average administrator may never want to explore that, but the specialist wants to tunnel into the details of what's happening in these systems. Got to admit if something went wrong and we still live in the physical world. So you really need both, but they need to be seamlessly integrated in a way that you need to be able to peel the onion and not see everything up front. You don't need all the visibility up front. The benefit of virtualization is that abstraction because it allows you to mask the complexity. So as you come down from the top down, you peel the onion, as you cross into these subsystems, they have very specific technologies, et cetera, underneath them and you need really both. But things are morphing too. I mean, you would just use the term mid-range. I don't even know what defines mid-range. But we just had NYSE on and I was surprised to learn that basically their whole back end is VNX. And I'd say, well, wait a minute, these are the most demanding applications. They're really good. Yeah, we love it. We eat that stuff up. Yeah. You think a bit more from an application perspective, right? I don't think people really blink. If you say to them, what applications are you willing to put on your mainframe? What application are you willing to put on your open system? What application are you willing to put on Windows and Linux? You intuitively know that. Because really it's not about performance. It's not really so much about price. It's really about the SLA, the quality of the overall service. So you still look at many enterprises and their most mission, mission, mission critical thing is still on the mainframe. And it has been in probably a long, long time. And as you come down that tier, so when you really approach it from what's the application requirements, then you get clarity on what the infrastructure is. And it just so happens that these mid-range systems, as we call them, have tremendous price performance. Eric's talked about what's interesting about these mid-range systems on, they're on a tear in terms of technology and innovation to produce IOPS and bandwidth that you could only get historically in a much bigger system. And by the same token, their capacity points are tremendous because they're very, very large now, right? But they don't have 27 controllers in SRDF and Time Finder and all these very broad bus services like you'd find in the mainframe infrastructure. So it's really kind of cool. And it's really extreme my own to the mix and it's like, sure, okay. And VMWare is a software mainframe. Exactly, exactly. So the beauty is there's lots of innovation. There's a lot of innovation. And you guys, you got a lot of resources. You got good cash in the bank. You got a good M&A team. You know, Tucci talks about it all the time. I was just down in Mexico four or five days ago giving the overall EMC story. And I was jazzed just going because you don't often tell the whole story. And then Joe talks about, we have two methods of innovation, right? One is organic and the other is inorganic. And so we build a lot of stuff. We spend 11% of our revenue in R&D and we continue to invest both organically and inorganically. And, you know, unlike many companies, we've kind of have the equation of how to take something inorganic and bring it into the family and really capitalize on it. A lot of companies struggle with that. A lot of companies will buy assets. Won't name anybody. He's up on the screen over there right now. But it takes him a decade to get benefits out of a technology acquisition. We just don't have it. Well, you know, I mean, in fairness, right? EMC wasn't always good at acquisitions, right? I can remember some early days, but you know, you got it right. I got it right now. The EMC's right up there. VMware, Data Domain, Icelon now. Extreme will be one of the next ones, which just, you know, we figured out and that's really a testament to Tucci and the strategy there. Yeah, I mean, it's obviously got a good process there. I mean, I've said a number of times on theCUBE. I mean, I put you guys up there with Oracle, with IBM, and there's a handful of companies that are really good at acquisitions and you're clearly one of them. Greatest acquisition in the history of the industry, of course, is VMware. Sitting right here, right. So, well, anyway, congratulations on all your successes. I really appreciate you guys coming on theCUBE. Thank you very much, Dave. I really appreciate it. Keep it right there. We'll be right back. SiliconANGLE.tv's continuous coverage. We're live from VMworld 2012. We're right back. Bye bye. Thank you.