 Live from the Mandalay Bay Convention Center in Las Vegas. It's theCUBE, covering VMworld 2016. Brought to you by VMware and its ecosystem sponsors. Now, here are your hosts, John Furrier and Stu Miniman. Hello everyone, welcome back day three of VMworld. We're here live in Las Vegas. This is theCUBE, SiliconANGLE Media's flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier, day three coverage of wall to wall three days of coverage at the hang space of the Mandalay Bay Convention Center. We're here at the SiliconANGLE Media team, research team from Wikibon. Stu Miniman, also my co-host. Peter Burris, co-hosting on the director's set and David Floyer, CTO at Wikibon. He's been scouring the landscape, sitting in briefings. Peter's been doing 50-50 briefings and hosting. Stu's been hosting all day. Of course, we've been banging out the coverage all three days. And so let's get into it guys. Analyst session here. I want to get the analyst take. Today's customer day on theCUBE. We have a lot of customers on talking about how they're using the technologies as VMware morphs into a broader, bigger company from a one product virtualization to a multitude of products and technologies. All trying to solve the complexity, as they say. That's in their DNAs, as VMware execs say. David, make sense of this. Are they doing a good job? Is the engine of innovation for VMware functioning? What's under the hood? Is it working? What's your thoughts? Okay, well, that's a broad range of questions. So is it working? Is it working? Yes. I thought this was the best VMware that VMware that I've been to. It was as much more confident of their role within the data center and within the service provider. Much, much more engaging, much more partnership orientated. And I thought that the start of it was the great confidence that they had in Michael Dell. He seems to be able, he seems to be signaling that he's gonna allow them all the latitude they need in the world to make partnerships within the Dell enterprise and also outside of it. I thought that confidence really came through as opposed to the handle on the throat that EMC often appeared to have. So that's a good thing. And I think it showed. You say EMC kind of had this invisible hand, the federation, back dooring into access. And you're saying you feel good about the sincerity of Michael Dell. That's right. If you look at VSAN, for example, VSAN six years, it should have been out three years ago in volume. And you felt that EMC was holding things back. Maybe for good business reasons, but holding things back. But we've just done our service and report on the VSAN itself and Nutanix and all the other players in this space. And VSAN have grown very aggressively. The quality of the product, the customers are really liking it. And they look ready to go. And obviously, following on from server virtualization comes the storage virtualization. Well, I want to just for the folks watching, just say that he mentioned the server sand report from Wikibon. Wikibon has put out the early definitive server sand statement, which by the way, wasn't really getting a lot of cheers. In some cases, a lot of people like this is right. A lot of people didn't agree with it. But now it was highlighted on stage. It's on wikibon.com. Go look for the server sand report. It's been updated. It's going viral. So in the show here, that report, congratulations guys. It's going viral. It's what everyone's talking about. And the VSAN success, I was talking to Parak Patel yesterday in the hallway. He's, you know, driving, it's exploding. So why is that going viral? Can you share just some color guys, Stu and David? Why is the server sand report that you guys did? Why was it so groundbreaking? Why is it going viral here at the show? Yeah, so just to give a little bit of context here. When I joined Wikibon six years ago, David gave me a reeducation on some of IT. I'd spent lots of years in enterprise IT, hardened everything, you know, really somewhat focused on hardware. And the joke we always have is hardware eventually fails, software eventually works. The software guys have a very different architecture. We understand that every component's going to fail. And we talk about hyperscale architectures. And when we first started seeing these early companies that were building what we now call Hyper Converge, we're like, wait, hyperscale meets enterprise. We were connecting the dots. When we put out that thesis in the marketplace, they're like, this is really enterprising. But your forecast, we don't believe this at all. And every year, the last three years, they say your data this year looks really good and we believe everything you said. But you're next year stuff. And after three years of putting this out now, they're like, wait, we keep agreeing with you every year. And we're going there. Maybe you guys really have something here. I mean, David, congratulations, it's been a pleasure to work with you on this stuff. The feedback at this show, not just from VMware, but all the other players in the market and the surrounding has been, you know, really vindicating after, you know, the hard work we've been putting in. So why is it going viral then? And what's the, so guys, explain, why is it going viral? Things just don't go viral because, you know, the track record is obviously there, so I agree. But why is it going viral? Everyone's talking about this one report. Right, fundamentally, the thesis is that storage is going from the sand itself closer to the server. And if you combine the two together, you get much, much better performance from the two. And it's also combined with flash. There's a lot more flash now. And the flash is taking away that ability to have things at distance. It's got to be close to the server. And what's happening within the server as a whole is now the network is the bottleneck. And now you're seeing the combination of much faster networks, people like Melanox, for example, much faster storage with the flash and attached and the processes themselves. The three together are making a much lower cost, higher performance. And if you look at people like Datacore, which are a small company, they've got benchmarks using this methodology, which are now five million IOPS. Absolutely out of this world, twice as fast as any of the original sand. And because of the configuration? It's all closer together. So another way of putting it is that it used to be that the latency off the disk and the subsystems associated with the disk was large enough that slower networks didn't get in the way. Now with flash, the latency out of flash is so low that you have to worry about the speed of light. David's right, it's pushing it back closer to the server without reducing the software interfaces and manageability. So it's not direct attached, but instead it's using software but pushing it slower so you can operate it faster. And also the application pressure too for real-time data, having stuff process faster, is that also contributing to this new kind of way of doing things? Yeah, it's a whole number of technologies. You've got erasure coding in there. People like Pivot3 have used erasure coding for years being the basis of this and again, lowering the cost. It's a continual application of a lot of different technologies and innovation that's happening in this area is mind-blowing. It's just fantastic. Yeah, so David, it reminds me a lot of what you're working on now is some of those same considerations of storage and networking and latency factor into IoT and edge computing and everything else. I know that's something you've been looking at at this show. Maybe you can give our audience an update as to that. Yeah, edges are funny words. Edge from an IT perspective is round the edge. You know, it's the factories, it's those funny people out there on the edge and headquarters is the center. From an IoT perspective, the edge is actually where the real value is being created, whether it's a nuclear power station or a factory or whatever it is. They take a different view of the world and there's a view that every, within a lot of people in IT, that this is the edge and we should have the data and put it all in the center. From an IoT perspective, they want to maximize the value that they get out of the data in real time, maximize that and hold that data probably much closer to the edge and then if people want it, come and get it from them. And they will be doing a lot of the analytics out there on the edge. So it's a different view of the world. We think that the edge is going to be an area of innovation within IT where it's got to be highly available, highly changeable from the outside, managed from the outside, but the edge is going to be an area where there's innovation, people putting out very, very large amounts of storage, et cetera, very close to the action itself. So Stu, we can update what you said. Hardware fails, software. Hardware eventually fails, software eventually works. Hardware eventually fails, software eventually wins, physics doesn't care. And we're doing a significant new piece of market research exactly on what David's talking about, to take a look at it. Anybody that wants to contribute, Peter at Silicon Angle, that's the direction we're taking. So the one quick comment that I would make is that I've never been to a VMware world before, but I've heard a lot about the challenges and tensions, many of which that David mentioned. One of the things that I see coming out of this, and it's going to be interesting to hear what the customers have to say, is the degree to which VMware becomes a salve, a glue, a way of bringing together the Dell and EMC ecosystems and helps use software to unify the whole thing. That's a great point. And I think one of the things that David said also sparks that conversation, I've been all show, I've been asking, always have these little questions, the puzzle pieces I'm trying to figure out. And one of them is the future of the ecosystem. And really what does VMware stand for? And so I've been kind of asking all the execs, these kind of pointed questions, but one of them I'll share, and I asked Gelsing in this and I asked some of the other ones. What's VMware trying to do? And actually the best answer came from Steve Herrod, who's no longer with VMware, but he was the CTO now as a VC. He said VMware's DNA is to solve complexity and make things simple. It was their original DNA of virtualization. And so what's interesting is I asked him, so what are they tackling now? What is the complexity that they're tackling? And how is that rendering itself in the customer environment? So the next series of questions goes to the customer. What's the complexity that you are trying to abstract away to enable faster acceleration of innovation? And that storage on the one hand and network on the other. I would add one more to that. And that is, we are going to enter into a period of uncertainty about inter-cloud communication. That we're going to see a pretty significant amount of money being spent on how do we ensure that clouds can work together? We saw some good data from VMware. It's actually a little bit less than I had thought, but the average shop, the average CIO is administering at least two to three relationships with cloud suppliers. Nothing wrong with that? They need alternative sources, but how you arbitrate the services out of those clouds so that you end up with more simplifications is going to be crucial. That's an interesting point. And I want to get you guys to comment on a point at the VC party last night, Lightspeed Ventures, all the VCs from Silicon Valley were there. And they were asking me, what's the hot startup is? And I go scratch my head going, you know, there's not a lot of big movement, but then I said, well, here's the areas that I think are going to be interesting. And the inter-clouding was interesting. And they, what is that? What's inter-clouding? Kind of, we're riffing on that term with Lutaker on theCUBE a couple of years ago, but it was what internetworking did during that explosion of TCPIP, birth, Cisco, 3Com. So this is a new area. So how does internetworking, if you use that analog, to inter-clouding? Is it the software? Is it the IoT? I'm not going to let David answer the question and get to the, but it's not unlike the experience that the industry went through when we went from mini-computer to client server. Fundamentally, it wasn't Intel and the PC that killed digital and DG and those companies. It was TCPIP killed the proprietary networks. The industry decided it was not going to bridge these networks, it was going to flatten them. And a similar type of thing is likely to happen in certain respects in some of this inter-cloud communication. But it all comes back to what David was talking about earlier. How do you handle storage? How do you handle network? How do you handle the CPU? TCPIP was open at birth, Cisco, basically at 3Com. It was TCP and fast Ethernet, man. So that really, what's open now, that's the similar disruptive enabler that could create the kind of innovation and wealth that that could- That is the big opportunity. To me, that's the big opportunity. Totally. At least in infrastructure. In infrastructure. Can VMware get into there? When you're looking more detail, you see that NSX is actually required for this inter-cloud communication. Are Azure and AWS going to allow that? I think there's an opportunity. Why is Matt Gelsing in that point? I'm like, how are you going to inter-cloud with Amazon? Oh, they have API, and like, you're going to sling APIs? That's your answer? Come on, Pat. Come on, go deeper than that. I mean, it's that interconnected tissue from the networking side. So, you know, it is another- Yeah, I think there's another answer as well. A little bit of a relationship there, just like Amazon with VPC is kind of reaching. That networking is going to be where we build some of those bridges and tunnels. I think there's another answer that may come out, which is equally important. And that is physical location. So if you have co-location with, for example, Equinix in that space, and you have all of the other cloud vendors in that same physical space, the people who are going to own that into communication are the people running that data center. The facility will matter. Facility mechanisms for allowing that. Yeah, well, physics matters. But just one thing, but when VMware bought Naysera, and Naysera can actually bridge between various hypervisors, and that's what they're doing. So let's not confuse physical networking with the virtual networking, which is really where VMware sits. But we got to be careful about bridge-based solutions, right? No question. Yeah, there's tunneling that we've been talking for decades about this. So everything we're talking about just proves that this is going to become a problem because we're trying to do, we're trying to make a lot of these systems, they're working, but performing some unnatural acts. And as a consequence over the course of the next few years, this is going to, David's right, this is going to be where it all happens- So what's the complexity that needs to be solved? That's the question. The complexity is, how do you get different cloud systems to communicate with each other at the data level? So that you don't have to segment your applications and your data? And I mean, step one is really, we're going to manage multiple pieces there because moving is, you know- Moving data is for the birds. I think I've heard many times today, it's more of just kind of the interconnecting tissue, not, you know, global V motion is, you know, maybe hyped a little bit too much over the years. I don't know that we're going to get, well, let's put it this way, we're not going to get a stateless TCP IP to solve the problem because what we're talking about involves a lot of state. So at the end of the day, this is going to be an area of a lot of innovation, a lot of invention, a lot of innovation. VMware is going to be in the mix. And I think that that's what we want to talk to customers about how is VMware going to be in the mix of solving these problems? The whole question of what open means, just because you use open source doesn't mean that you are open. There are new lock-in mechanisms evolving that don't require any proprietary code, but could have some nice sticky factors like data. So it's going to be very interesting next couple of years, guys. I mean, Microsoft's got the big move to make right now. I mean, don't you think? I mean, what they do now is to be very telling what Microsoft does with Azure Cloud is going to be really, really telling. 100%. Okay, guys, analysts breaking it down right there. They're breaking down the opportunities for a creation of innovation and wealth and jobs in this area for startups and the big companies also where VMware could go. This is theCUBE breaking it down. We're going to hear from customers today, day three. I'm John Furrier with Stu Miniman, Peter Barris and David Floyer here on theCUBE. We'll be right back with more. You're watching theCUBE.