 We're here with Kirk Brezniker from HP, BCS, Business Critical Systems, chief technologist. Welcome to theCUBE. Thank you very much. I'm also joined with my co-host, guest analyst from Wikibond.org, Dave Cahill. Dave was an analyst on Wall Street, as runs a firm called Diligence Tech Advisors. One smart guy knows this stuff about the business. He's into storage, cloud, infrastructure, and we're going to demystify the notion of high end systems are going away. With every new era of seismic shift or inflection point, as they say, it's always the death of something, right? Kirk, so it's the death of the mainframe years ago? That never happened, just got smaller. So you're in the high performance system, so first question is, tell us, you overlook the hardware, look at the software, HP's a systems vendor, they bought Compact, they've been in the server business, HP UX, they're the only mini computer actually that's still alive that I can tell, besides IBM, you got software, you got hardware, you got Compact, you've got DEC, Digital Equipment Corporation, as part of that, kind of integrated in. Got a lot of stuff in there in the toolkit. Where are we today with the high end systems and demystify the notion that those systems are not happening right now, they're going away. Sure, so one of the things that really happened, you gave the whole pedigree of all the pieces that we brought together, and one of the things that's really great about being in business critical systems is that I think my first trip to Houston to meet my Compact compatriots was about a month after the HP merger, we got in a table, not much bigger than this one, we had the nonstop fault tolerant team, we had the business critical system team, we had the ProLiant team, we all came together as one engineering committee, and we were able to pull together all those pieces, something that we had never had before, which was a huge volume business, we had this deep technology IP well that we could draw from, not just from BCS, but nonstop and the alpha systems, the deck systems as well, we were able to pull all that together, and I think if you went out in the floor today into the Converged Infrastructure, you'd really see the fruits of that labor, that labor of love of pulling all these technology pieces, and for us as high-end technologists, it was really opening up a new avenue to express the engineering because of that volume business and adding that in as well. So for us, I'm not sure how the other guys can do it, but it's great for me to be able to draw upon a server business and a server supply chain that produces a server, several times a minute, we're pumping out a new server, and I can use those same components in my high-end systems as well. So regardless of whether or not you want to massively scale out, or whether you want to retain a massive scale-up environment, I'm able to do it and able to do it for less money to acquire, but less money to own, and for me, selfishly, less money to develop because I have this huge technology base to build upon. The computing industry has always been about parallel processing, a lot of high-end computer tech distributed computing adds to that, and you have a well of knowledge, and it sounds like you guys were kids in the candy store putting that all together, but we're seeing some of that same concepts push down. Vertica, they talk about massive parallel processing in the big data analytics side that they do in clean sheet of paper, so these same trends are coming into other parts of the computer business now, which has gotten smaller and smaller down to a smartphone, metal fertilization, solid-state memory, these are not new concepts in theory to the higher-end systems, so you guys are in a unique position where you have the technology at the high-end, and you have a fully-built company that sells the other stuff, so share with us some of that cross-pollinization that you guys have worked on and some of those fruits, as you mentioned. So I think an excellent example is the technology that has now become the HP Cloud System. When we looked at how would we want to manage virtual and physical servers, independent of the choices our customers would make to optimize their data center for these new emerging technologies, we actually didn't have to look too far because we already had technology to manage virtual and physical platforms, it was just in our Unix environment, so we took that technology, migrated over, that became part, when coupled up with new technologies like Virtual Connect, that became the Cloud System and the Cloud Map. Now we've sort of pulled that full circle, as we saw demonstrated this week, being able to pull together Unix platforms with HPUX, Windows, Linux platforms on X86, whether they're virtual platforms, whether they're physical platforms, all combined in one single Cloud Map. Basic technology we brought over from the Unix side, coupled with really emerging technologies on the X86 side, and come together as one coherent offering. So from an R&D perspective, Kirk, do you see the, I mean, the world is increasingly modular and increasingly X86, but yet they're borrowing all of the design principles that have been at the high end forever. So is it now, it's almost R&D leverage, like we've done this before, guys, let us help show you the way, how to get there, right? I think that's a good way to capture it, you know, when we work on our platforms like the DL980, the top end of the ProLiant platform, we realize when we reach a scope and a scale with a Xeon platform of that capacity, that it was an opportune time for us to pull technology over. So we took technology from our Superdome 2, our SX3000 scalable chip set, and refactored that into the Prima architecture that's now driving our DL980. So it's an opportune time for us to cross-pollinate some of this technology. And I think when we see some of our competitors, you know, they tend to silo. You have a mainframe, or you have a UNIX platform, or you have an X86 platform, and they don't really allow that kind of customer use across those. I think that they're missing an opportunity. It's one that I think we're exploiting to the benefit of our customers. The ability to have, you know, common components, power supplies, IO cards, memory dims, up and down an entire range of products. How far along are you there? I mean, is that real today? Oh, absolutely. If we go out on the floor today and go to the Converged Infrastructure Pavilion, you'll see a Superdome 2, you'll see integrity blades, you'll see Proliant blades, and things as simple as a power supply. Pop it out of a Proliant rack server, pop it into a Superdome 2. And so, you know, the modular X86 will borrow from the high end a lot. What are you guys borrowing from the X86 guys? What are they teaching you about compute server processing that you can incorporate into your systems the other way? So for us, it is things like that base technology. I think that also, you know, one of the things I admired immediately upon the merger with HP and Compact was all of the electromechanical, the physical user interface pieces that really differentiated Compact Proliant servers in the X86 marketplace. You know, we get the advantage now of all of those pieces. Technologies like the onboard administrator, which we have now pulled into the Superdome 2, enhanced with our additional partitioning capabilities with our things like our Superdome 2 analysis engine, but those manageability pieces, the whole inside suite, tremendous software assets, hardware assets brought back and forth across the product lines. Right, and so when you think about cloud, how do you guys as a business unit think about that? Because it's traditionally on premise, high end, very sensitive data sets. What's the opportunity to infuse yourselves in the cloud discussion, whether it be private, whether it be public. I mean, how do you guys, when you're sitting around the table talking about strategy and direction, how does cloud become part of that discussion? Well, I think the first thing is that even though we traditionally talk to customers and they are, as you characterize them, very sensitive to data location, very sensitive to being able to demonstrate where data is living throughout the life cycle of an application, at the same time, they want to restructure their IT departments to move from the silo pillar, put a server in, give it an application, have someone tied to that application, have someone tied to that piece of equipment for the next 10 years. They want to have the same agility, and that's really what we're able to do because we've leveraged that common technology. I can still have a cloud map with an H2X piece, with a Linux or Windows on x86 piece. Now, chances are that our customers are going to want to maintain that kind of capability in a private cloud on premise, but we still want to afford them all the modern technology, all the insight control, all that dynamic capability, even if it's going to be on their own premise because it establishes that IT as a service-oriented relationship with the line of business. So we don't want them to have to maintain separate, different ways to procure and deliver and associate IT with lines of business. We want it all to be that service-driven economy between an IT provider internal who is then making those decisions to flex external as required, but also can demonstrate that kind of data integrity that is required for mission-critical applications. So does cloud map span the hardware? I mean, is it hardware agnostic in that sense? I mean, this app catalog and idea of being able to, you know, service catalog and agility, et cetera, is that cross-hardware platform? Absolutely. So if we were to pull up the cloud map tool, I can pull on a server and I can drop on HPOX and I can drop on HPOX applications in the same way I can drop on integrity virtual machines just I would have done with x86 servers, with Windows, with Linux, with any of the x86 virtualization platform. I can drag and drop and pull and network and storage architectures on those and then hit the button and have it published in the same short number of minutes as I would have had for x86. Right, so it's hardware agnostic on the cloud map side. You got common components up and down the lineup. Where are we going? What are you thinking about now? When you think about what's next and directionally, futures, technology, what direction do you think this takes and incorporating it across the line because there is a commonality there on the application and the component side. So directionally, what's next? What do we want to talk about next year? So I think that if we look forward to next year, we're really sort of just broaching now this ability to transform the relationship between IT provider and the line of business and then I think we'll see this accelerate as we change from the traditional model of I deploy services, I deploy servers, I deploy applications on them and they stay static. So I think that over the next couple of years you'll see more IT organizations embrace this service model and then that just sets them up to make these kinds of underlying technology changes independent of the relationship with the line of business. Now I think as we look into what are those technology changes, I think that changes around the storage hierarchy, the incorporation of high performance, high capacity IO targets, the change of operating systems and how they will optimize around things like solid state storage. All these will begin to factor in the change to not just traditional OLTP or business intelligence, but the combination of big data on unstructured data that then couples to equivalent structured data queries. All these things are going to factor into how we deploy this infrastructure but as we're positioning our customers to take this first step into the cloud system and into these technologies that reshape this relationship between IT and line of business. It positions them well to incorporate these kinds of novel technologies much faster than they would have traditionally done if they had to stand up a new silo with a new technology and try and find a solution that the line of business is looking for. Right, because things like flash, fusion, PCI, you know, PCI based storage, et cetera. I mean, that enables completely different mindset at the application layer than it might have previously, right? And I mean, I guess it's a mindset thing where application developers have to approach this completely differently but it's enabling them to do things that they couldn't have done before. That's correct. For some of them, it's not just the throughput, it's really the latency. If I can look at my business every hour, it's going to change if I can look at it every minute. And now we're talking about it is every second to really real time. And how is the line of business going to change its behavior if I can offer them something that is really going to give them that kind of telemetry, that kind of finger on the pulse of the business, how would they react differently? Is your life easier or harder with 3-par? I think it's very exciting for us. That's a great question. You know, as far as looking to having that storage which is optimized for these kinds of new deliveries, I think we're just really excited about the inclusion of the 3-par technology into enterprise service storage and networking. So I guess my question would be, you know, in terms of the use cases on the customer side, you're looking at a lot of different views, landscapes, what are some of the hot trends that you're, that's the thermals that you guys are riding that you're basing a lot of the future kind of tech decisions on some of the tech? I'll see a lot of tech at the high end, huge customer base, big productions, critical systems. You can't really tinker with it. But for the most part, you do have to match some of the architectural shifts that are happening and they are now, version networking, mobile, et cetera. So what are the big trends that you're riding or seeing that you're shaping your future? So I think we already talked about one, which is, how is the storage hierarchy going to change when I move from traditional media to incorporating high performance, solid state media? And that's where I think, you know, the addition of things like 3-par with autonomic tiering is a way for us to begin to bring that in, to begin to see how that's going to affect things. For us, as far as the advanced technologies that some of my team are working on, we also look to things like photonic communication. Again, trying to change the ratios of compute to memory, change the latency equations. And it's really not talking about what has happened today, which is, oh, here's a piece of copper cable and it's more inexpensive for me to deliver this with a laser at both ends and a piece of glass in between. What we're really talking about is looking at entire ranges of this infrastructure and saying, can I take a whole section out? You know, what would it be if I could erase the walls between these servers in their massive scale out? I wouldn't have a scale out model, I wouldn't have a scale up model, I'd have pools of resources that I could begin to manipulate very quickly. So I think that those are some of the things that we find really exciting. The intersection of big data with solid state storage, with very high radius photonic communication that's also very low power. And how is that going to affect the software infrastructure that people look to, the operating system that people look to? When we talk about these large pools of resources, we sort of come full circle too, back to wanting to have very robust technologies and software infrastructures that are able to manage and manipulate those. And it just so happens that we have, you know- Core systems. How do you do that? Lots of really great experience in managing huge resources that run indefinitely. And in the case of things like our non-stop environments, massive scale out, massive fault tolerance as well. Okay, we're here with Kurt Brezniker. Thank you for coming on theCUBE. It's you guys run a high end, the business critical systems. No matter how the architecture changes, no matter what happens in the future, it all comes back to the customer. They still need those core systems. They got to be high performance, high securely, highly available. That game is not changing. The players may change in terms of the hardware and approach, but the tech is ultimately the same. Thanks so much for coming on theCUBE. Really appreciate it. Thank you.