 15 to 20 years ago, we saw function move out of the host, storage function specifically, into what became the storage network. And for good reason. We needed to share data, we needed to protect data, we needed to get data off site. And moving that data away from the host made a lot of sense. In the past five years, we're starting to see that trend reverse. And there are a number of factors driving that. We're seeing a massive increase in processing power, decreasing costs in both processing power and memory, and also the disruption of flash. So you're starting to see Das as a new mechanism. And I'm here with Stu Miniman, who just released today a new definition of server sand. We're going to talk a little bit about that today. And we're also here with CJ DeSise. CJ is the president of the Emerging Technologies Division at EMC, a group that is set up to specifically identify disruptions in the industry and then bring them to market. CJ and Stu, welcome to theCUBE. Thank you, thank you. David really appreciate it. So happy to be here. So let's start. There's some news today actually, Stu. You know, we're talking about all these trends and shifting sands in the industry. Well, David Goulden was finally, it was announced that he's the president and CEO of the EMC II group up here of Pat Gelsinger now and Paul Moritz, which of course we knew all along. We've been saying. Yeah, David, it's really interesting. The nuance on that, of course, is that Joe Tucci still is chairman and CEO of EMC the Federation and EMC squared the entity itself. But now Goulden is a peer with Pat and Paul running all of the kind of storage infrastructure, the IIG group and all those pieces under there. So CJ, your new boss got promoted. So in fact, you got promoted. So congratulations. Congratulations. So let's talk a little bit about the emerging technology products division and what the scope is of that. And obviously your role and why you, why did the EMC choose you and why did you choose EMC? That's a great question. So I joined in September. My background in 90s, I was with Oracle Corporation and one of the things that I did at Oracle besides doing Oracle applications and so on was we were the first group to launch software as a service at Oracle. This was in year 2000. This was an RMCRM online in that era. And then I was with Symantec. Originally I'd end the security division, which was called Endpoint and Mobility Division. And then after that, my last job with Symantec was running Veritas. So everything from VXFS to NetBackup, BackupExec, Storage Foundation, and other technologies archiving in the cloud and so on. So the reason I joined EMC, Dave, is because they have healthy paranoia about disruptive technologies. They want to bring disruptive technologies in-house or invest in disruptive technologies, give it proper resources, whether it's related to R&D or sales and marketing, to make sure that rather than anybody else disrupting EMC, EMC itself can disrupt when there is a market demand for disruptive technologies. Well, I think that's an interesting point too. And still, we've been observing this market for a long time. Being from the East Coast and watching all the mini-computer companies get crushed and of course, Joe Tucci sort of came out of that world. It is very much a tune to that. I think EMC and others have a healthy paranoia. I think that the legacy of today's leaders is largely going to be determined by how much they don't have their head in the sands and how much they can sort of eat their own. And we've seen it in a number of instances with EMC. I mean, we're a close example. Yeah, Dave, so I mean, no secret. I spent 10 years at EMC and one of the Bibles working at EMC was only the Power Noir that survives by Andy Grove. So EMC was always looking out for that disruption and both through internal R&D and acquisitions, they've done a number of moves to move beyond just being a hardware company. They expanded their software and their services division and it bought a number of product lines to kind of bolster the portfolio. Okay, so CJ, the role of the emerging technology products division is essentially to identify disruptive technologies, is that right? Talk about that role a little bit and how that all fits into the EMC and the Federation. Right, so first of all, at EMC-II level, which is the EMC storage and infrastructure division, we are focused on looking at the technologies that are disrupting the market. So let me spend a few seconds talking about it. So one of the first assets that is part of this division is Extreme IO. As you know, there are many all-flash arrays that are out there in the market. EMC has an architecturally superior product that was designed from ground up for flash. So that's one product line that is part of the division. The other one is Scale IO and you're talking about things moving to the host and with my experience in Veritas and so on, we are definitely seeing the trend that there is definitely unused capacity attached to the server. So can we use DAAS with manageability of sand? Can we use, can we help our service providers who have scale out infrastructures? They are under cost pressures and a solution that is software-based and attached to server and can leverage flash or commodity hard disk and so on is game-changing, is very disruptive. And there are a few other initiatives happening within Emerging Division, which we'll talk about at EMC World, not ready to talk about that yet. Okay, but in concept, should we look at the division as both a place where disruptive acquisitions like Extreme IO come in and Scale IO, and is it also a skunkworks? Of sorts. Skunkworks is a tough word, but yes, advanced research is what I would call it there. So we are doing- We're very colloquial here sometimes, that's what you're talking about. So we are doing advanced research in certain areas. We are exploring partnership where it makes sense for EMC and we are also constantly looking at disruptive technologies that fit in with overall EMC's vision. So in theory, that is the idea behind this division. So let's talk about, before we get into the service end, let's talk about flash as a disruption. I mean, you guys, well, we wrote a piece, I wrote a piece, actually EMC lands a haymaker, when EMC, a long time ago, when EMC put the first flash device into Symetrix. It was 2008. Yeah, we look back at that now and go, okay, that was sort of not the future, but at the time it was an indicator of the future. Yeah, 2008, that's right. So flash as a disruptive technology, give us your thoughts on flash as a disruptive technology. Other than the fact that it's a persistent medium, what makes it so disruptive? So I think one of the things that we have seen is that the demand from our customers is very high when it comes to performance expectation, low latency, consistent high performance at low latency. So like you discussed with VMAX, the journey started out in 2008, then followed by VNX, and both VMAX and VNX can go from zero to 100% flash in the configuration today. Extreme I.O. specifically was designed from ground up to leverage the power of flash. And what we are seeing with Extreme I.O. specifically in the use cases is when you have virtual desktops, right? As applications are moving more towards the data center and adoption for virtual desktops is happening, the performance given by Extreme I.O. on an all-flash array is just blazing performance. And then you have hyperscale and other types of customers who are saying we want to attach flash to the server, we want a solution that can scale out and so on. So we are seeing flash being attached, driven through the traditional send, as well as also as part of the server infrastructure and that is disrupting the market because the demand for performance, whether it's virtual desktop or virtualized servers or even OLTP workloads where you have high expectation when you're doing reads for the data analytics and so on is very high. So you mentioned the hyperscale, you've got the true scale out architecture and then the traditional stuff, you got a stack that's robust. And now a lot of people say, okay, new is going to kill old and it never happens that way. But eventually do you see these two worlds coming together and how does your division sort of cross-pollinate technologies? You share with your colleagues, borrow from your colleagues, how does that all work inside of EMC? Great, so EMC definitely one of the reasons Dave I joined is because it has strong lineage in send. And when we are trying to even for extreme IO where we can leverage the scale with engineers and understanding the storage infrastructure, architecture. So for example, we have done integration with bunch of EMC products such as V-Plex and so on with extreme IO. So we do leverage that scale quite a bit. And we do that leverage the scale with sales and marketing efforts with our core Salesforce and so on to make sure that the division and the products that come out of the division, one fits in with the overall EMC strategy and two can scale, no pun intended, can scale out as we scale the businesses. So that's number one. And then number two is what we also find is I am talking to all of my peers and all the engineers in different organizations to make sure that we are not, we are leveraging the experience they have had, the rich experience EMC had. And if we can do things differently, then we'll do that. So for example, on scale IO, it's a great piece of technology is server attached software storage solution. Now, Viper is our overall software defined storage messaging. So we are gonna integrate with Viper when it comes to both the control pane and the data pane. Now that sand background that EMC has the perfect lead-ins due to the whole server sand discussion. Now, I mean, I love the fact that SiliconANGLE Wikibon, we tend to be ahead of the curve in terms of things that we track. I mean, we were on cloud in 2006, 2007. We were early on Hadoop. We've been writing about Bitcoin. You scratch your head about Bitcoin. Now Bitcoin's everywhere. Server sand is something that people have talked about, Stu, and there really hasn't been a clear definition of server sand. Like there wasn't with software-led infrastructure, which became the software-defined data center. But so you took it on yourself and some of your colleagues, David Floyer, to define server sand, as they say, a term that's been bandied about. But so what is server sand and how does it relate to this discussion? Sure, Dave. And in a lot of ways, this is really an extension of what we've been tracking at Wikibon for a while. One of the big trends we've been following since I came on almost four years ago is how does hyperscale impact the enterprise? So scale out architectures, how Flash is involved in things and really having software be the layer that controls everything rather than hardware really differentiating too much is a major piece, as well as convergence, which is an area that I've spent a lot of time on. How are the compute and the storage assets going together? So server sand really sits right at that intersection. So it is a converged infrastructure, but it's built out of compute nodes that have local storage and the software layer itself can manage all of those storage functionalities. I think back to CJ, you worked for Symantec. Symantec and Veritas for years were one of the toughest competitors for EMC because they could offer very inexpensive solutions that provided a lot of the functionality that EMC was trying to do with their storage arrays. And now with the advancement of the software functionality, you're really seeing that blurring between what is a hardware solution, what is a software solution, as well as that server and storage line has really blurred a lot. So the hyper-convergence players kind of fit into this discussion, what EMC is doing with Scale.io, fits square in here, VMware's vSAN, and a number of others. So it's an exciting new space. And as Dave you said, over the next few years, we expect many applications are going to fit into this environment and it will directly impact the tradition. So just to clarify, so server sand is essentially a pool of direct-detached storage devices that communicate, how do they communicate? Is it a high-speed internet, whatever, or an infinite internet or whatever it is. And the point of control is the software, right? Okay, so that's exciting, because again, years ago, we were talking about who's going to be the next veritas of the industry. There's a software opportunity enabled by the performance of microprocessors, the decline in cost of microprocessors in memory, to the point where you can actually, without the trade-offs, have software control the storage pool. So that leads to Scale.io. I wonder if we could talk about that a little bit. Where does it fit in the portfolio? Talk about the acquisition and where you guys see that going. So Scale.io became part of EMC family in the early fall timeframe. So this was around August, September of 2012. And the reason we were excited about Scale.io is because one, like we talked about, for service providers, you see that they're constantly under cost pressures. They want to use the commodity hardware where it makes sense. And Scale.io is block solution that works closely with server and of course blurring the line between the server and storage, like you were talking about earlier, between host and storage, but provides you full manageability, parallel IO, massive performance, and can scale up to thousands of nodes. And one of the advantages, Dave, of Scale.io is it also works with heterogeneous infrastructure. So whether you have multiple hypervisor from VMware, Microsoft, and so on. We are also doing integration with OpenStack. We are also making sure it can work with bare metal, flash, and magnetic disks. So it is truly a heterogeneous solution. So that's the enterprise use case. Sorry, service provider use case. And then on the enterprise side, we are also excited that when enterprises want to leverage desk with manageability of sand, then Scale.io is a perfect solution, which was your earlier point on the warehouse side. Well, we were talking about the hyperscale market before, and when you dig into what's going on in hyperscale, they're not putting in big sand infrastructures. They're using this direct attach, you know, approach. But of course, the challenge is the average enterprise doesn't have, you know, a zillion engineers running around like Google or Amazon. So is that the role that EMC is playing, essentially bringing that hyperscale class of product to both the enterprise and the service provider? That's exactly, you got it correct. So when you look at it, the service providers, when they want to leverage the manageability and all the features that you expect in traditional sand, such as thin provisioning and so on, this is provided by Scale.io today. So that's the big advantage of Scale.io in the service provider market. And on the enterprise side, if you want to run DevTest, if you are not comfortable moving the data to cloud and you want to run quickly and use your desk capacity or on commodity storage, you can do with Scale.io. And the third thing I'm excited about is Viper integration that will integrate with Viper also to ensure that from control plane of Viper, you can manage Scale.io and Scale.io and provide data services. So let's talk a little bit more about and unpack the sort of the applications and workloads. I know, Stu, what you found in your initial research, but where does server sand fit? And then CJ, I wonder if you could just sort of comment. Yeah, so Dave, definitely there is a little bit of that gap. This scale out architecture is built for, will be great for some modern applications. So desktop virtualization, it can definitely fit there. Some big data applications depending on how they're configured can fit there. And but it's not today built for necessarily the same kind of mission critical databases there, but that's where over time we expect server sand to expand with more functionality and more. Is that consistency, CJ, with what you guys are saying? Yeah, we are definitely seeing that when you are, so for example, you have commodity hardware and you have bunch of blades from a server provider and you are running an application that your sole purpose is to scale it out at an efficient cost, this is a perfect solution because this will give you the manageability that you require that traditionally you expected in sand, but of course at the same time you're gonna be cautious. If you're a service provider, cost is the driver, you're gonna go for it. Enterprises, they will start with something like Dev and test to what he was saying and make sense. What about converged infrastructure? Where does that fit? One of the things that surprised me a little bit that we were recently at Amazon Reinvent is to hear Amazon talk about how they're developing more customized solutions. You see the OCP initiative that came out of Facebook with sort of horses for courses configurations. How do you see converged infrastructure fitting in to your whole space? So of course as you are aware, very well that VCE was first attempt across the federation and in partnership with Cisco to go after the compute networking and storage and VCE has been received very well by our joint customers. We are also looking at few other options to leverage the technologies we have such as the Scale ION, the V-Sense of the world to ensure that we continue to be player in a converged market. But that is definitely, I call it in layman's term data centered in a box, it is definitely a preferred option. You don't have to configure it as much and you can scale out as much as you want. So you're definitely seeing demand for converged infrastructure. So you're wondering if you could just give us, I mean, you're not going to give us hard metrics but if you can give us just a sense as to how it's going in the vision. I mean, to do something like this in a large organization is not trivial. So you've successfully put the organization together. You've made a couple of acquisitions. How's the uptake of customers? I know at EMC world, there was a lot of enthusiasm or certainly around what they were showing with extreme I.O. But maybe you can give us an update just in general terms, market momentum, things that you're seeing, enthusiasm that you have and then we'll wrap. So, Dave, good question. So what we have seen is customers are definitely excited about these two technologies that we have. So first is from an all flash array conversation when the performance demands are very high, a lot of our customers are coming to us more like a pool model and asking us about an all flash array or an existing array, what performance at what cost we can give it to them. So definitely a lot of momentum on extreme I.O. business already, we G8 the product on November 14th. And so it has been on market for six weeks. Prior to that for a few months it was in directed availability. So what we have seen is now customers are able to install and are pretty happy with the performance they are seeing in the VDI use case or virtualized servers and so on. With scale I.O. we have seen again demands on both sides. One is service providers are definitely having conversations with us, some of the large names. And also on the enterprise side when they are trying to leverage desk capacity they're definitely exploring that. So we have a great momentum behind us. The company has been very supportive to ensure that these technologies are successful. And so far I'm very positive about the future of these technologies. So my last question is what should we be paying attention to? What are your sort of goals over the next six, nine months? Milestones that observers should be watching as sort of indicators of progress. So we are going to of course evolve the roadmap when it comes to say something like extreme I.O. and continue to add features. As I said the architecture was designed from ground up to leverage flash power. So we'll continue to add features over time and do multiple releases. When it comes to scale I.O. we also have another release planned in the Q2, Q3 timeframe. So there will be features we'll add there too. And then there are other technologies that are being worked on in our division and hopefully we can announce that at EMC. All right CJ, well thanks very much for coming to our studios in Massachusetts. Appreciate you making the trip up here and thank you Stu for the good work you did. You released this morning and keep it right there. We'll be tracking this issue and the evolution of ServerSAN and EMC's emerging products division. This is theCUBE, thanks for watching everybody. We'll see you next time. Thank you.