 Live from Las Vegas, Nevada, it's theCUBE, covering EMC World 2015. Brought to you by EMC Brocade and VCE. Welcome back to EMC World, everybody. I'm Dave Vellante. Nick Kirsch is here. He's the Vice President CTO of EMC's Emerging Technology Division. Nick, welcome to theCUBE. Thanks, it's great to be here, Dave. E-T-D Day, today, day three. That's right. You guys are going to be excited. Great keynotes going on, and you're the CTO, presiding over the technology direction of all of it. It's got to be exciting times for you. It's very exciting. I've been waiting for this day all during the show. So, Icelon is the crown jewel of the Emerging Technologies Group, and you've got another shiny, bunch of shiny toys that are coming in. That's right. Talk about the portfolio and how you view your challenges, your agenda from an architecture standpoint. Absolutely. If you look at the Emerging Technology Division, it's all about finding these new technology elements, going out and meeting those bleeding-edge customers, and either incubating the technology with them, or incubating the go-to market for EMC, incubating the way that we look at our partners. So, there's lots of different challenges here. Some of those are technical challenges. Some of those are business challenges. You saw we went free and frictionless with ScaleIO. We open-source Copperhead. So, these are significant business changes that EMC has to do to really become viable and proactive in some of these emerging markets. Yeah, so, we heard from CJ this morning a lot of talk about open-source. So, it's kind of new, was sort of interesting, some challenges early on. Talk about EMC's relatively new entrance into open-source, and what that means for the company and your customers. Certainly, so from an Icelandic standpoint, we've been participating in open-source for a long time. We're based on FreeBSD, so I've got a lot of personal history with that, but this is the first time that we've really been able to open-source an entire product. So, rather than leveraging open-source and contributing back to it, this is the entire product is open-source, and we want to foster active community and development. And if you look at Viper, it's a great platform for it because we need that multi-vendor support. We need those customers stepping up and saying, hey, these are the high-priority items for me to manage a variety of different storage subsystems. I want to be able to contribute my own ideas to this platform, and so that's a great place for us to enter this. Here's our product, it's open-source. Come, contribute with us, build it together. So, let's start with Isilon. I mean, you get a lot of traction with the whole notion of Data Lake. Early on, people said, oh, big data, it's commodity, scale-out hardware, just off-the-shelf stuff, but then the Hadoop met the enterprise, and the enterprise wanted resilience, and they wanted this sort of enterprise capabilities, and everybody started talking about, we're going to make Hadoop enterprise-grade. That whole market's kind of fell into your lap, didn't it? It sure did. Maybe talk about that a little bit, and then met them you're seeing there. Absolutely, I think of you, when I talked to Hortonworks and I talked to Cloudera, their vision for HDFS is really Isilon 10 years from now. And so we're in there where these customers, they want to see a large-scale distributed file system. They want to be able to apply next-generation analytical techniques, whether those are Hadoop-based techniques, or whether they're things like Apache Spark, but they want to combine it as part of their workload with their NFS-based techniques or their Windows-based techniques. They don't necessarily, they can't uplift their entire process and move it to a net new sort of analytical technique. So, Isilon steps in. We allow them to bridge this gap by giving them all the enterprise resilience and features that they really want today. Yes, so the base architecture is there, but you had to do a lot of work to enable HDFS, right? What specifically did you have to do to sort of connect into that Hadoop world? Yeah, in some ways it's simpler than you would think because we were already a large-scale distributed file system, so we had to speak the name node and the data node protocols on the wire, so the Hadoop clients have no idea that they're talking to an Isilon versus a commodity HDFS array. And it was literally that simple because all the metadata and file systems were already distributed, were already distributed storage. Okay, so that explains kind of why you guys have had early momentum there and kind of taken the lead. All right, let's talk about software only. You guys announced Caspian this week. What's that all about? Yeah, so Caspian really, I think the way to look at this is this is bringing simplified next-generation infrastructure to these enterprises. There's really two varieties that we're focused on out of the gate. One is open-stack focused, one is big data focused. And it's this idea that, hey, I could take some next-generation techniques whether I'm all Hadoop, whether I'm focused on Cassandra, whether I'm building applications that are containerized, and I don't have to go and build this as a part-piece sort of function. I can deploy scalable appliances and get all the ease of management to do with it. Okay, so give us some more details. You're talking about things like, well, open-sourcing the Viper Controller. That's right. Now, it's interesting, there is a technology component there, but there's also a business component there. Maybe talk about those two different vectors. What does that all mean for EMC? Yeah, in some ways, the technology component might be the easy thing. We understand how to open-source. We see lots of examples of open-source, but do you think about it? EMC internally doesn't even have a shared code base among our various groups and departments. So one of the biggest benefits I see at a Copperhead is now we've got this open-source model that EMC teams can contribute, rather than sort of going to a Viper team and saying, hey, can you build this for me? The ISLON team or the DSSD team can take it upon themselves to contribute changes and grow this inside the EMC portfolio. And that same advantage is to our customers as well. So as a CTO, I let people always talk about, Joe Chuchi says it's better to have overlaps than gaps. So you've got a lot of different products. It's always a challenge to make those things work together. From your perspective, how do you attack that challenge? Do you try to optimize on the product and its business value sort of for that use case? Or do you sort of pull back and say, okay, how do we integrate these pieces? I mean, you've got a lot of diverse pieces in the portfolio. That's right, that's right. I think in some ways, ETD is a simpler portfolio because we're chasing growth opportunities and those growth opportunities are fairly well differentiated. But from a basic level, it really is about the use case and the workload for the customer. What is the best fit? You want to optimize for what the customer is going to experience over the near term, but you also have to have a strategic view. So to give you some concrete examples, you could look at something like ECS or Project Caspian and say, boy, it'd be great to have a robust file services in this platform. Or you could look at an ISLON and say, boy, it'd be great to have geo-distributed objects in this platform. But when you look at the customer use cases, there isn't actually a lot of overlap between how they're using these systems. And so we don't tend to build for that. We probably would make those same choices if we only had that product as our portfolio. But it also allows us to think reasonably about the long term, where do we see these big growth areas coming? I mean, you're essentially an arms dealer. And so- Absolutely. And you mentioned ECS, you take ECS as a component into the hybrid cloud solution. So you're basically selling into the solutions group, whether it's internally within EMC or externally to your customer base. Does that notion of cloud solution minimize the degree to which you have to integrate across the portfolio? It's certainly the case that customers are certainly coming to us with a solution's first approach. And so when we come and say, hey, I'm trying to solve this particular business challenge and we give them a set of EMC products that together solve that challenge, then it's not about the individual product differentiation in any way. Okay, so what's the minute reaction? I mean, you're talking to customers at the show, right? Are they getting it? Are they excited? Are they confused? Are they lining up to get the products in and what are they doing with it? Yeah, I think in general, there's been some good excitement. I think you hear some of the questions you asked here. I, hey, wow, EMC's got a lot of products. How do I figure this out? Some customers are like, yes, I can see how I'm going to use all these things. Other customers step back and say, wait, Stream.io, DSSD, Scale.io, which one am I picking here? So there's definitely some education there because we're at the bleeding edge of some of these workloads. Well, it's interesting. I mean, you know, everybody talks about companies in transition, the innovator's dilemma. It seems like EMC said, okay, let's organize to address that. There's the platform three guys, you know, go. How are you seeing the take up within the customer base, particularly around sort of digitizing their businesses? Is it more they're using your products as a bridge or are they diving in to the future? Yeah, I think it's fair to say that it differs widely depending on the customer. There's certainly customers that have early adopters found uses of EMC technology and dived right in. I think Isilon is a great example, as you say, it's a large and growing business and most of the places where Isilon got in first was customers that were leading the adoption curve in some sort of digitization of their business. We see that Isilon typically lends itself to an object store approach over time. And so it's customers at one of the large file system over time, they move that architecture to a large object store. And so ECS and other conferred sort of object opportunities come in nicely into that RAM. You see the same thing with DSSD where they might have taken an Isilon as a large distributed file system from a performance standpoint, but that next level of performance is an in-memory approach and DSSD starts to lend itself. So we actually see a lot of sort of continuity in the customer approaches between these products. What are you seeing far as open stack uptake? Open stack, I'd say, it's funny, I think we're sort of in the trough of disillusionment for open stack to some extent. I'd say that there was a lot more open stack conversation last year, a lot more customers who were determined to go open stack. Now you see a split, you see customers that say, hey, I'm going to keep my ESX environment around for a long time. Because a lot of times open stack is primarily being used to replace ESX or to augment ESX. So I see more customer confidence in maintaining an ESX environment for a longer period of time and sort of experimenting, wait a second, what is the right place for open stack? It's not really a perfect replacement for my virtualized environment. Where can I use it more effectively? And Nick, when you think about the whole software defined meme, what's driving customers towards software defined? Is it an economic model? Is it a flexibility model? What's the motivation? I think it's both of those. So if you look at sort of large customers that have large control over their data centers or are making data center level optimizations. It is about flexibility and control. It's about optimizing the individuals that are working on those components. Facebook, Microsoft, these are great examples where they're doing data center level optimization. And so by standardizing on hardware components in that entire data center, they're driving at lots of costs in the construction and operation of the data center. And they want to take advantage of the IP assets. In other places, it's pure flexibility that customers want to move up the chipset faster. They want to be able to have some pricing flexibility when it comes to the hardware components. I think even more interesting is customers who just want broader flexibility. Can I deploy this in a virtualized context? Can I burst into the cloud? If I'm going to standardize on this kind of technology and I'm going to train all my employees to use it, I'm going to become an expert on it. I want to know that it's flexible and that I'm not forced to only buy an appliance in the future. When a trend like software defined gets defined and marketed, everybody sort of hops on board. You saw it with cloud. You certainly saw it with big data. You're seeing it now with software defined. There's a lot of software defined washing going on. What makes software defined, software defined? Is it the API, the ability to programmatically provision? What's your view of that? Yeah, in some ways, I think it's many of those things. I think at the base essence, it is a software engineering strategy that has designed this in absence of any kind of tie to a hardware environment. And usually we don't talk about this as software defined, but scalability is the key component of that. So if you're going to build a true software defined infrastructure, if you're only scaling to a small number of nodes, what's the point of software defining it? This is really about large scale. It's about starting out with thousands of nodes as you start a target deployment and seeing that customers can easily grow there. And so you can have a lot of strong hardware ties. Yeah, well, and the programmatic provisioning of things and scaling is fundamental to that. That's right. All right, listen, we're out of time, Nick. Thanks very much for coming on theCUBE. Really appreciate your insights and good luck going forward. My pleasure, Dave. Thanks for having me. You're welcome. Keep it right there, everybody, we'll be back. This is Dave Vellante. This is theCUBE, we're live at EMC World 2015. Right back.