 Live from Dublin, Ireland, it's theCUBE, covering Hadoop Summit Europe 2016, brought to you by Hortonworks. Now your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live in Dublin, Ireland for theCUBE, Silicon Angles flagship program, where we go out to the events and extract the signal of noise. I'm John Furrier, my co-host Dave Vellante, our guest guest is Steve Sarfsfield, marketing manager of Big Data HPE in Boston. HPE, part of Vertica team, welcome to theCUBE. John, it is great to be here. I am so happy to be here. I've seen so many of your videos. I feel like I finally made it. Thanks for inviting me. Well, say hello to Colin for us. He's been pretty much on theCUBE every time we have an event and move us to Big Data event you guys are having coming up in the summer in Boston at the Weston. Yeah, we have a big event every year where we bring in our customers and some of our prospects, analysts, and we talk about Big Data. We talk about very specific, you know, Vertica functions, functions that are within our software platform. So, you know, it has been a great event, very popular among our customers. And yeah, I'm looking forward to it. It's a really good event. You've had some cool outside speakers. Billy Bean spoke one year. Of course theCUBE is there. And we were there at the first one. John and I have actually done some stuff on stage. It's been kind of a rogue event, Dave. I mean, we went far in the beginning. Colin and his team had a great Chris Ellen put that together with the team. You guys are great. But you had the DevOps crowd early and that conference was kind of rogue in the HP event. Now it's kind of formalized. But really what you guys have done was you've shown the developer early days of Big Data and hardware, high performance, compute, storage, and software. Heavy hitters. Absolutely, and you know, and the customers really like to talk to developers and that's what they get at that conference. It's not a lot of marketing fluff. It's guys like Chuck Baer and Ben Van Diver who are talking about the internals of the product. So we're real happy to put it on and the customers just love it. So you guys with that have attracted some real significant players. We call them tech athletes. People who are doing the DevOps, web scale, hyperscale, from Facebook to all the people doing the cutting edge, large scale systems, networks in the cloud and on-prem with Vertica. Dave's always said that Vertica's been one of the best acquisition in the history of HP in terms of contribution and price to value. But you guys had that unique positioning with what I call the Ferrari of storage with the MPP database that came out of Vertica. But now you guys are winning these massive deals with some really serious players. But it's not a one-trick pony in terms of the architecture of your customers. It's sitting in that. Can you explain the dynamic of how you guys are winning and why? Yeah, absolutely. We've been in this business for 10 years now ourselves. So back in the day, Mike Stonebreaker wrote this white paper on C-store, on column store. And from that he spawned a lot of businesses, right? But Vertica was one of them. And so our engineering staff has been developing this big data platform, the MPP platform over those 10 years. And really the MPP platform is kind of two parts, right? One part one is how do you store the data so that it's effective and fast. And when you pull a server blade out, it doesn't crash and you don't lose data. So how do you make it bulletproof? How do you make it enterprise grade? The second part of it is the analytical functions, right? How do you take all the SQL functions, Python, R? How do you perform analytics on all that data in a very fast way? So there's kind of two parts of the MPP platform and we've done a lot of development and a lot of work really fleshing that out over the years. So we've talked to a number of your customers, particularly at that event. Yes has been on, Zynga, Facebook has been on and it's consistent that the value proposition is raw performance. I mean they could do things that they couldn't do before. Does that continue? You were sort of hinting into it there. Does that continue to be the value proposition? Or is there in the differentiation or is it evolving and what is it evolving to? Yeah, I mean that really is a very big differentiator. Just being able to run your queries, even in your data warehouse, talking about a query that may have taken you eight hours before now can run in eight seconds. That's a really big differentiator, right? When you can sort of have a better conversation with the data. You can talk to the data, it answers back and then obviously when you get those answers you have more questions. So that's kind of the way that Verca helps with that. I think there are other things though. I think some of the shortcomings of other big data platforms are things like concurrency. How can I have hundreds of users, thousands of users hitting up against that data and really performing analytics against that, against the data at the same time. It's tough on some of the other platforms and it's really easy on Verca. The management aspect, that's another part of it. How do I manage my cluster? How do I manage my queries? And that's an important part that we've built out as well. So being able to manage, being able to do the analytics fast, being enterprise-grade is really where we're at. I'm going to ask you my security question. I've been asking all our guests here. What's changed in security, specifically in the context of the way CIO should be communicating to the board of directors? What's changing? Well, I think one of the promises of Hadoop is that you can take data and you can put it in this big data lake and the nice thing is you can have access to that data. You can use it for whatever you want. You can do some discovery. You can try to understand it. But it is just a big data lake. And so some of the strategies that folks start with is just setting up a firewall and saying, okay, we're not going to let you in and take a look at the data lake and issue of credentials. But there's so much more to it than that, right? So HPE acquired a company called Voltage a few months ago and that has been a great acquisition for us. So now we have HPE Secure Data, which can help with securing data when it's at rest, being encrypted and stored, when it's at motion, so when it's traveling through your network, and also when it's actually in use. That can be a fault as well with the data. So data in motion, data at rest, and data in use encryption is where we're at. Talk about the dynamic that's big trend in the industry, separating the compute layer from the storage layer. A lot of conversation is going around, where's the glue layer? You got a data layer. And it's evolving part of the stack that's really coming important. And what's the dynamics around it? What's your thoughts on this? I mean, I think kind of the old way of doing business, HPE sold a ton of those ProLiant DL380P servers, right? And it's pretty much a static server where you have sort of a static compute model and static storage modeling. So the old model was, hey, I want to put more compute in, I just buy some more DL380s and put them in. I think people want more now. They want more out of their platform. And so we now have the Apollo platform and our big data reference architectures that allow you to sort of separate the storage and compute. So if you need to add more storage, you can do that with the DL380s. If you need to add more compute, you can do it. So think about it, when we have an ETL job or we're doing AI, that's more compute. You need more compute for those types of jobs. If you're archiving data, then you need more storage. And so being able to customize what you buy is really key with the big data reference part. You know, I noticed that Antonio Neary, great CUBE alumni, Antonio would love you. It's been participating in HPE, participating in Facebook's OpenCompute, or OpenCube, probably not Facebook's, but OpenCompute project where Facebook's a big part of that. And Facebook had their big announcement yesterday with Mark Zuckerberg on stage, put in a 10-year roadmap. They showed a camera that they're going to be open sourcing through OpenCompute. This brings the trend of hardware and software, and you guys have it both, right? So this is the equation that customers are dealing with. Do you guys talk about things like that where you maybe, because you're going to separate the control layer and the active layers, you got software as a key part of that. You guys have these conversations? Yeah, I mean, I think you bring up a couple of questions there. So first, how do you handle the video and the audio? And so we have a part of it called Haven on Demand that can help you kind of understand facial recognition, do voice and video and audio. So that's part of the big data platform that we can offer. I don't think many people know about that, but that's an important part of it. This high-performance workloads is a mess. High-performance workloads, absolutely. And right now it's cloud-based, but I think people will want to bring that in. But the other part of the question is about sort of separating the compute and storage layer. So we work with partners like Hortonworks, for example, in making sure that our hardware and software works within the yarn system. That's a great way to allocate resources by working with yarn. So you can set yarn labels, for example, to allocate different compute and storage within the hardware. And also you can use yarn to configure how much resources you want to use for queries on the analytical, on our Vertica analytical platform. So there's really a good integration with other vendors is important. You mentioned Haven, and what generally sounded like AI, at least touching upon it. Obviously, the IBMs make a lot of noise with Watson. Facebook yesterday at F8 was talking a lot about sort of its machine learning and AI. Google obviously has some chops in that space. Is that your AI play and how does it fit into the whole big data space? Yeah, it really is. And I think what customers want to do is they want to look at three different types of data. They have the business data, which is your CRM and ERP data, and that's growing by leaps and bounds. You have the Internet of Things, which has some structure, but the problem there is it's just huge amounts of data that you have to handle, generally. And the third type of data that you might want to look at is voice, video, audio, human-generated data. And so really that's what the Haven platform is all about, sort of taking that human element, that human data, and understanding that and performing analytics on it, and that's what we've done. So in thinking about the challenges that customers are having today in big data, we heard in the audience this morning, most of the people actually were using Hadoop, doing so-called big data. I think maybe 10% of the audience was just putting their toe in the water. So what are you seeing as the big challenges today? It's been historically, wow, this stuff is so complicated. You know, I can't get an ROI. Now that's not been as big an issue for Vertica, because Vertica's been solving the sort of slow EDW problem. Absolutely. Okay, but then that dragged you into the Hadoop mess. So what is the big problem today? Is it still complexity? How is that evolving, and how are you helping solve it? Well, you know, I think companies want to adopt the latest technology to solve the latest problems that they have within the industry. So they want to adopt Hadoop. They want to adopt Cloud. They want to do on-premise databases. And so what our challenge is, and the things that we're building out for current and future versions of Vertica, really talk to being able to access data no matter where it sits. So you have an orc file, you have a parquet file, you want to access that and make that part of your analytics. You can do that with Vertica. You want to store it in Vertica on the Cloud, or you want to use a virtualized environment. You can also do that with environment. So kind of, you know, making sure that you can apply analytics anywhere where the data sits is a key part for us. You don't want to have to move the data if you don't need to. And so applying the analytics to where it sits is really the best solution. So we've talked about Cloud a few times here. We've had Bill Hillfond, we've had Bobby Patrick on talking about your Cloud strategy. We've been following it closely. We understand sort of HP's strategy of not trying to compete directly with the public Cloud offering. We think that's the right call. Having said that, you know, some of your competitors would say, oh, we don't have a public Cloud, you know, HPE as a result is not a viable option. What do you say to that? Well, you know, we work with some vendors that are in the market now. So we are, we have a pretty good relationship with Microsoft and the Azure system. We also have a pretty good relationship with Amazon and AWS. So Vertica can run in those environments. They can run virtualized in those environments. And it's really, you know, kind of using, it's using analytics where our customers are asking for it. Yeah, so we think that Cloud is much more of an operating model than it is a destination. It's not a place to put your data in your apps. It's the way in which you affect your operations and automate and take out non-differentiated heavy lifting and shift it toward application development and business value. Can you do, first of all, do you agree with that? And can you deliver that type of experience to customers where the operating model is essentially seamless, independent of location? Yeah, and you know, I agree with that 100%. A lot of our customers are kind of pulling us into this area where we originally didn't conceive Vertica to go and that is sort of virtualized environments, right? Setting up virtual servers so that they can set up their own private clouds. And Vertica now has many customers who are using virtualized environments. So it is an area where we're being pulled into and we're doing some development around that. I think that's where it's going. So I mean, it sounds like you're sort of opportunistic with respect to what if the customer wants we can deliver that. How about flipping that? What kind of proactive discussions are you having with customers in terms of the strategy as it relates to their operating model, specifically in the context of data and where Vertica plays? Yeah, I mean, you mentioned earlier the big data events that we have. So we have customer focused events. We also have a advisory board that comes and meets once a quarter. So they help us with our roadmap and where we want to go. What they've been asking us for is things like, I want to do a specific type of analysis. So we have this thing, this new feature in the product called live aggregate projections. I don't think I need to go into that, but it's a feature that was completely spurred on by some of the customer advisory board meetings that we've had. So analytics are part of it. And also sort of the whole idea that you want to take data and apply analytics no matter where it sits, that's actually spawned by our customers as well. So we're taking out those ideas and really helping our customers drive toward that. Steve, what are we expecting to see at HP Discover? I know that's a little bit out a little bit, but it's right around the corner. It's coming fast. Yeah. Vertica in front and center. How's the big data team? What's the software team doing? Give us an update. Yeah, I mean. So how's that being sorted out? So you will see what's happening with the latest version of Vertica. We have some new versions coming out, so we'll be talking about those features. And we have sort of Haven on-demand platform that will be part of it as well. I know I have a couple of talks there I'm giving where I'm talking about the concept of having multiple databases within an organization. So why you need both an operational and analytical system. So we'll be talking about all sorts of great Vertica, big data platform things. Make sure you come to Discover. And you got the big data event going on in August. I think it's the last week of August, is it? Yeah, absolutely. How about, yeah, that's right. Very last week of August, first week of September. Yeah, head to head with VMworld. How about life as an enterprise company? What's that's been like? It's been what, five months in now, four months in? It's been incredible. And I think it has sort of garnered a new life out of HPE, HPE, the software side of it and the big data platform side of it. So I think when we all found out about it as employees of Hewlett Packard that the company was going to split, it was kind of a shock and we were expecting pain. The split has been really smooth as an employee. It's been great. Morale is up big time. Morale is up big time. And I think also, we're more focused on the mission. The mission to provide the big data platform, to provide hardware, software, services, everything you need in order to. There's no place to hide anymore. There's no place to hide. HPE's out there in the open now and there's the enterprise group. Yeah, yeah. There it is. Plus you're balanced. We can't wait for HPE to discover. HPE Discover, I should say. Plus your balance sheet is a lot better than it used to be. I can't comment on that, but yeah. No, I can. It's very rude for Meg. We're waiting to see Meg on theCUBE, Meg Whitman, hopefully have her on. But we've rooting for, I think, a focused organization, Antinua Neary, great executive, all CUBE alums. All the HPE, pretty much CUBE alums, except for Meg Whitman. But I know you can't comment specifically, but don't you think that, I mean, from a morale standpoint, that gives employees much more flexibility. Meg, you know, a couple of years ago said, flat out, we're not making acquisitions until we clean up the debt. Well, that's been cleared up. From HPE's perspective, you don't have a lot of debt. You know, just seeing how smoothly everything went in terms of the split and just seeing how there is a new environment, a new feeling to this new company, we're ready to take it on. We're ready to take on the big data world. Steve, thanks for taking the time to come on theCUBE. Appreciate it, great to see you. I appreciate your coming on. Thanks, John. Appreciate it. Thanks for your insight. HPE, big data, Vertica, check it out. And they're having a big data event. All their top developers really showcase customers. These are the elite forces in their world. It's in Boston, check it out. Big data, I don't know if it's a big data event. I think it's called, it's in the Western and Boston at the end of August. This is theCUBE, live in Dublin, Ireland. We'll be right back after this short break.