 We're back here live in HP Discover 2012 in Frankfurt, Germany. I'm John Furrier, this is theCUBE. This is our flagship program. We go out to the events and extract the signal from the noise and at HP Discover, a lot of great news. HP is about big data and software led infrastructure. That is the focus here and everything wraps around that. Really beautifully, software, solutions, services, everything that the customers are building upon. HP's addressing it here and that's a real positive sign for HP and we're covering it live. I'm John Furrier, the founder of siliconengle.com. I'm joined with my co-host. I'm Dave Blomethick, and software led storage. So we'll go look at that. We're here with Paul Miller of HP. He's the vice president of Converged Application Systems. Welcome. Welcome, thank you. Great to be here. So we took a lot of buzz. We took the liberty of saying it's all about big data and software led infrastructure, which actually is our favorite areas that we report on. But I think in general, although it's not your official marketing, it pretty much summarizes where the market is. Would you agree? Yeah, I agree. And what my team does, I'm uniquely positioned in HP. I sit within the enterprise group. My team is the guys who marry the software, the big data software and the applications with all the hardware. So I think I'm in a great position with HP. Get to play in all the worlds, right? The hardware and the software world. So big data seems to be the big buzzword but that wraps around basically the application market. So when Dave Donacelli was on earlier, we talked about kind of the theme of inflection points. So you had mainframe and client server change the game significantly, then it went to best of breed. Now we are in a whole nother transformation mode that's going to look a lot like mainframe and client server kind of wrapped up really quickly. And that's really around applications and compute. Same paradigm, compute and applications. It's just that the definition of those two are changing. So I wanted to ask you on that point, what is the app side of the market and how has the big data been into it? Obviously today the low hanging fruit is analytics. Clearly no brain, everyone wants analytics. So what do you see around the big data from your perspective on those two points? Compute and computing and applications. Right, so if you look at big data and computing, right? Everyone's talking about the growth of big data and how it manifests itself. But the next big trend is going to be around the variety and then what unique applications can customers build on top of their foundation? So if you look at Hadoop, Hadoop is just merely a big data lake that you pour all your data into. And what I think the next generation is going to be is building as Hadoop and then you wrap tools around it, management tools, tools to help query, and then building the vertical applications on top of it that allow the marketing organizations, the finance organizations, the business people to actually interface with the data in a way they haven't without being a database administrator. You know in the past, that's what you had to do. The next layer of applications is going to be probing on that big data pool without you knowing that you're a big data analytics expert. You know I asked Dave Donatelli a question. Dave, how do you market your value proposition because we're in context to software defined networking, et cetera, et cetera. And he had a great answer and this is really my question to you. We got to build the products first because you have to sell the products and wrap the services around it. So the question to you is, with Hadoop and big data, there's a huge pressure from the customer and demand from the customer base to actually have product. So can you share with the folks, what is the core products that you're developing and what are you dialing up now to be in that roadmap? Okay, great. One of the big problems with Hadoop is it's brand new. Not a lot of tools wrapped around it. So if you're used to rolling out traditional databases, all the tools you're used to don't exist in that world. So what we've done with the appliance we announced this morning is wrap it with rich tools to enable you to one, stand up in a Duke cluster really fast. And with this, what we're announcing is this new dashboard. We call it the app manager for Hadoop. On one screen, it gives you the entire cluster infrastructure, you know, most of these big Hadoop farms are hundreds of thousands of nodes. How do you manage? How do you get your arms around that? This dashboard gives you one shot to look at all your clusters, all your software enablement, all your environment in one shot. You can do push button updates of your firmware, push button deployments, you need to scale. Gives you a view, an integrated view, looking at your cluster to understand where's the hotspots? Do I need to move workloads around? Do I need to balance off? Change my configuration. It really gives customers that one stop dashboard to manage your entire Hadoop environment. So customers don't need to be afraid of Hadoop. Now they can embrace it and deploy it and have high service levels on Hadoop like they're used to on an old traditional legacy database. And that's what's critical to environments as they move Hadoop from being a science fair project into the mainstream and into mission critical because now business leaders are going to be dependent on that data. So is that really the differentiation strategy is you don't need a thousand map-reduced programmers to deal with this, we make it easy? Make it easy, make it easy to manage the setup and then with tools like Autonomy and Vertica, you can actually probe the data in multiple different ways, right? Business intelligence used to be all about building out the map schemas, right? All the schemas, understanding the data and then what questions to ask. With Hadoop plus the combination of Vertica and Autonomy the business user can start asking the questions without having to understand what's in the data set because Hadoop and we make it easy to propagate that out. It's really a change in the marketplace of how analytics and big data are merging together like no one's seen before. All right, John, let me turn it back to you. I got my one question in. Yeah. Now you have more, I didn't know long was the question. So what you're saying is obviously there's a lot of build out and the tools are critical. You need picks and shovels to get a job done, right? So check. So my next question is Autonomy. Take us through the integration of Autonomy. Obviously we're seeing Autonomy sprinkled all around on the different announcements. David Scott yesterday did a great press announcement, press conference where there's an amazing set of technology and improvements. One of them was he had some Autonomy sprinkled in there. It was awesome. Big data storage kind of go together. Tell us about what's new with you with Autonomy specifically on the product side. Okay, so one we announced an e-discovery appliance as well. So our e-discovery product is typically bought by the legal department of compliance, it's not an IT cell. And what we wanted to do was develop a system that the IT department and the business department could deploy simply. It's a modular approach. So it can click into the Hadoop database where they store all their documents and provide almost instantaneous value to the end customer. The other thing you'll see us do is take Autonomy to the cloud. Another thing my team does is build out the cloud maps which are the best practices and the way you deploy applications in the cloud. Automated provisioning, the configuration to set up all from a single screen where you can download this and push button way and provision automatically. We're doing that with the Autonomy discovery set, e-discovery as well as some of the other Autonomy assets bringing them and making part of our N10 cloud strategy. So it's really exciting to take and I think take Autonomy into the mainstream of some of our selling motions. So what about this trend toward bringing SQL and NoSQL together, bringing real time to big data? What do you make of that? How will HP capitalize on that? What's the strategy there? Yeah, so I think what we have is with the Vertica as was the traditional SQL and Autonomy which is NoSQL, the best of both worlds. And by combining those with rich services and management across both of those, we can provide customers access to 100% of their data, be able to migrate and probe on 100% of their data that no one else can. So we really see those two worlds and the data set of the future being the mix of SQL and NoSQL. So, I mean there's a lot of talk about connectors. You guys are one of the first, I think Vertica actually was the first to do a connector to Hadoop. And I think actually initiated it on its own. Yeah. And then sort of brought it to the marketplace because nobody's really doing it then. There's a lot of talk about sort of going beyond connectors. What are your thoughts on that? Are we going to see like super connectors? Are we going to see more native integration? What's HP's angle on that? I think some of the native integration and running are native within Vertica and as well as building out arrests around the ecosystem. Integrating Vertica into Hive and Autonomy into Hive and really taking beyond just the traditional Hadoop HDFS. But now integration with the rest of the tools that are coming out within that Hadoop ecosystem, if you will. You know, Hive is one example that I think is everyone's probably familiar with. You know, there's upcoming different initiatives. You look at all the different open source initiatives that plug into Hadoop. That's to me the next holy grail, if you will, to monetize and make Hadoop real. Is that what customers are asking your help with? Are they asking your help in, well, making Hadoop more reliable? I presume that's one of them. But are they asking for other assistance and well, how do we actually make money at this? Yeah, so surprisingly. What's the answer? Every customer I've been with, they have a Hadoop instance. Sometimes the CIO doesn't even know about it, right? It's sitting, it's kind of like the early days of Linux. It's there and it's growing and it's migrating. But then you'd have a lot of customers, even some of your most high tech online customers are saying, wow, we got all this data. What the heck are we going to do with it? So a lot of what we're doing is linking the use cases of this customer did this with their data, et cetera. And we've developed a set of practices around Hadoop and big data for customers to say, one, what's all the data you want to put into Hadoop? Where's it coming from? Both the data within your house and data external to your corporation. And then secondly, now that you have all it, what are the use cases that are going to provide value out of it? I would say most customers are at the stage that they're starting to pour data into the Hadoop cluster and only have one or two use cases that they have rolled out within an environment. But that's where I think we can add a lot of value by just sharing among the community about the things we're doing with other customers and bringing that all together. Well, it seems like you've started to get more aggressive with regard to really reaching out and integrating the ecosystem. Is that true? Is that deliberate? Is it sort of just the natural fact that HP's so big you're everywhere? Well, one, it's I think a natural fact that we're so big, but two, you start to see HP embracing open source in the community around open source and ecosystem. If you see what we're doing with cloud and our open stack, we see open stack and then the Hadoop environment built on top of that as a real great combination for customers because when we can do integrate open stack and Hadoop together as our open stack strategy rolls out next year, that means cloud and big data on premise as well as in the public cloud. And I think that's part of our core DNA to be open and drive that strategy for customers. So you mentioned open stack, but you guys are serious about open stack. And again, you were early on in that movement. Now pretty much everybody's in. But I can't tell whether they're in because they're checking the box. Hey, we're open, we're open, we're open. My question is, is open stack ready for prime time? So HP's spending a lot of time hardening open stack. So I would say it depends, right? There are multiple different versions of open stack, quite not ready for prime time. You'll see it's going to be ready for prime time very, very soon. Okay, I wonder if you could Paul, if you could clarify something else for me. I listened to Meg talk about the strategy and she talks about information optimization. Is that another term for big data? Is it a superset of big data? Can you just sort of describe it as- It's really a superset of big data, right? And you start looking at information optimization. It's optimizing not only the big data, but your traditional data, how you access the data. So we're not only trying to go after the new emerging piece, but also the classic environment of what's traditional data is all about. And then how does that data manifest itself across everything? Obviously big data's white hot today, but information optimization covers broader span. Well, it seems like the big data's the tail that's wagging the information dog right now. Do you see that mix flipping? In other words, you got sort of the traditional database world is really where a lot of the value is today. Do you see that changing over the next 10 years where the unstructured stuff actually becomes the lightning rod of the organization? Or are those two worlds going to stay separate? Are they going to collide? What do you see? I believe those two worlds are going to combine and converge, right? So I talked to folks about converged information and converged big data, right? Because today they are separate, right? And the magic comes when you can start to converge the two worlds. You know, if you start looking at unstructured data for customers talking about the medical industry, right? Everyone's had an MRI, right? X-rays, et cetera, that become part of your legacy and become part of your history. Two generations from now, if that data's not, today we throw that data away. My view is we can have the technology to keep that data and make that data context rich so that your child and your child's child can use that to prevent disease in the future. And that's where, you know, a simple example of the two worlds can collide and provide a totally radical different way that the medical community services you by not only knowing your grandmother's history by what you checked off on a piece of paper, but by actually being able to see the data and make analysis on it. Right, so we need to remember your grandmother's history, right? Exactly. Paul, I want to ask you a question about kind of the HP big data umbrella. Obviously predictive analytics, sentiment analysis. I mean, we've been bullish obviously on big data, as you know. Share with the folks out there something about HP that they might not know about around big data. Because this show, it's pretty apparent that big data really is going to be a part of HP's future growth. Because it just, it is what it is. It touches everything that you guys are building out into. Mobile, touching into converged infrastructure, touching into services and software, all the above, HP will be touching big data in the marketplace, whatever that is, it's everything. So, explain to the folks out there about what is big data around HP? What's the, what's going on? How do you structure it? What's the organization? What are you guys delivering? Obviously Vertica's only one element of it. Now you've got autonomy. How do you wrap that up? Yeah, so I think that's one of the big mysteries about it, right? It's so pervasive across everything we do. And for customers to have a successful big data strategy, it just can't be standing up on a dupe cluster. You know, once you stand up on a dupe cluster, you need the rich tools like you talked about, Vertica and Tommy, to extract value. But then you still need to back up and have the compliancy. So you start putting on the compliance piece of it. A lot of what my team is doing is bolting the security piece of that, understanding what we can do from a security standpoint. And then building out a set of infrastructure tools that enable people to actually deploy big data in a very, very efficient way. Right, if you look at, you know, trying to deploy a traditional database for big data, you're not going to get there financially. You need new tools. You talk to Dave Donatelli, store once. They've been able to archive off that big data today that can't be kept resident and hot in a very, very efficient way. We're talking about, you know, our new hardware system that was announced two weeks ago, the SL1450. The ability to have an optimized compute storage node that can help customers scale. You'll see us talk about, that we announced our strategy around Moonshot. You'll see us work and have cartridges designed for big data. So whether you're looking at- It's native. It's all- It's native. It's actually so, it's part of the DNA of almost everything we do from services to hardware design to the application of data sets. It can't be separated. I think that's the bumper sticker. Big data is native in all of HP's products and services going forward. Okay, Paul Miller, BP Converged Application Systems for the Enterprise Group. The Enterprise is booming. No recession in the Enterprise. A lot of buildout going on with big data. A lot of tools you guys are developing for customers. I just had Tom on earlier about the service aside. Big data opens up a conversation that expands into a lot of different things from business re-engineering to technical deployment. So congratulations. You guys got a nice road ahead of you. And we're looking forward to following you. So this is siliconangle.com's theCUBE. We'll be right back with our next guest after this short break.