 Okay, we're back live here at SiliconANGLE.tv. It's theCUBE, our flagship telecast. We go out to the events and talk to the smartest people who can find CEOs, executives, CTOs, entrepreneurs, partners, whoever we can find, extract the signal from those here that with you. I'm John Furrier, the founder of SiliconANGLE.com and I'm joined by my co-host. I'm Dave Vellante of Wikibon.org and we're here with Kirk Brezniker who is the chief technologist of the BCS group at HP. BCS is the organization that focuses on mission critical, the largest, most important applications, running financial transactions and very high availability, high performance apps. Kirk, welcome to theCUBE. Thank you very much. So, HP Discover, we're here. You guys got a big presence. I was walking around your area, your pavilion earlier. A lot of partners are there. Mission critical, what's happening in that business? What's changing? You know, for us, mission critical is a segment that we have been leaders for decades in and that's in our non-stop environment, our HVUX environment, our open VMS environments and for us, what we announced, starting actually at Discovery in Vienna last November was our project Odyssey, which is, for us, our future definition, redefinition of what mission critical means. Now, when we look at mission critical, we actually are expecting that as a market that will grow. Now, it will change over time. If we look at today, mission critical is primarily epic risk-unit systems, mainframe systems and there's some x86. Undoubtedly, we'll go around the floor, talk to people here and they'll say, yes, I have, I'm running some mission critical, some mission critical applications on x86. What we expect over the next several years is that the overall mission critical market will, first off, it will expand. It's going to get bigger and I think we only have to look at the number of companies that just have to have mobile access 24 by seven, all around the world, all around the year and know that that is going to have to increase. Now, inside of that increasing mission critical bubble of addressable market, we expect the overall risk epic, that Unix market to stay relatively the same. It's fairly flat, it's sort of plateaued and it will continue on. What we do expect, however, is that customers are going to want more and more x86 to provide mission critical. They want this because they want to align architectures, they want to align software and they want to align experiences up with the volume economics that they have now grown accustomed to on x86. At the same time, they don't want to lose any of those characteristics that have gotten them through for decades and for us, that's what Project Odyssey is. Looking at all that IP we have in HPUX, in non-stop, in an open VMS and understanding how we can transition that and compliment that existing Unix environment with an x86 environment that has both the same infrastructure characteristics of availability, reliability, dependability, as well as mission critical services. When we have customers who have mission critical platforms, it's not just about the reliability of the platform, it's that when the rare occasion that something does go wrong, that all the information about what happened is captured, is delivered to an HP laboratory, root cause analysis found, corrective action comes back and that whole cycle happens in a matter of hours, not days, not weeks, not months. And so we're working very hard to design the services on top of mission critical Linux, on top of mission critical Windows to go along and compliment the hardware. A lot of people, you know, I've worked, back when I worked at HP, I just saw a colleague, I used to work with 27 years ago at HP, at HPUX, and HP's made some bold moves, PA risk architecture, and we're not going to, for the folks out there, we're not going to comment on the whole chip processor or Oracle lawsuit, it's really not relevant, it's very political, and it's really not a focus here, and I'm not going to ask you any questions there. But I do want to ask you that HP has made these bold moves, and they've had some, you know, some architectures, and they've worked with Intel, some stuff flew, some stuff did great. We were at NAB, we saw some real powerful workstations at NAB, so talk about that convergence between really high end, powerful, Unix-based, BSD, whatever you got going on the Unix side, and X86's, out of the X86's is that wind tail platform and the UX is the power, used to be the power workstations, talk about that convergence specifically, and what in market today are you guys focused in on? So if we start at the base, you know, we do have that very strong collaborative relationship with Intel, right, so that's both on the Intanium side and on the Xeon side, and we are working with them to educate them and to partner with them so that we have the appropriate foundational technologies in the microprocessor. I think the layer above that is in our traditional platforms, we have HVUX, we have non-stop, we have VMS, technology for scalability, reliability, the containment and processing of errors and continuous operation. Now if we look over in the equivalent on the X86 side, you know, that is where we're going to layer, the operating systems are going to be Linux and they're going to be Windows, right, so we have a great ongoing relationship with Microsoft looking at our deep IP and setting what we can pull over and partner with them and drive into both Microsoft server as well as things like Microsoft SQL database. On the Linux side, you know, that's an open development community and so we are going to be partnering there and driving some of our IP into there, things like scalability, how large a Linux environment can be, as well as the error containment, the ability to have a very robust and reliable Linux environment. Now some of that we'll be driving into, for example, kernel.org, we'll be working with specific distributions, Red Hat is our premier partner on Project Odyssey, but we'll also be driving some of the technology instead underneath the operating systems, so underneath Linux and Windows, so some of that IP that we have in HVUX and VMS and non-stop will end up being embedded in system level software inside the platform underneath the operating systems. So for us, we're going to be driving our IP below into the operating systems. What's that going to enable? Because that's an interesting move, because you're essentially dropping it underneath, you're abstracting away from the OS. What is that going to enable better automation at the hardware level, the chip level? Better things like containment of errors. So we are going, we have a very large scalable systems. We've announced both our Dragonhawk Superdome 2 with a Xeon platform. By driving some of our IP that used to be in the operating systems in HVUX, for example, and driving it underneath the Linux or Windows, it means that people can use the shrink wrap versions of these operating systems, and things like advanced error analysis and detection and correction can happen independent of the operating system that's chosen. So you can still use your normal operating system, the same one you're using on the industry-leading Proline, but it will just behave better. It will work better when it's on a platform like a Dragonhawk with things like our advanced error analysis engine, an embedded platform that's constantly monitoring and maintaining the self-health of the platform. And all this will happen independent of the operating system and underneath it. Things like partitioning, these are elements that will happen underneath the operating system. So you can still use a consistent operating environment from the smallest, thinnest Proline play that you use all the way up through a Dragonhawk. So, Kirk, you talked about the growth of the market, the mission critical market, which I buy, by the way. Certain segments of that market might not be growing, but overall, clearly the demand for high availability, mission critical, tighter RPO, RTO environments is growing. No question about that. Some people might say, okay, so we talked about Odyssey and X86 meets mission critical. Some people might say, well, why is HP forking its X86 effort? Why not put all the wood behind one arrow, so to speak? How do you respond to that? What's the strategy there? Well, I think that there is one offering in one converged infrastructure. So when we look at a Dragonhawk, our Superdom 2 for integrity as well, you're looking at a platform that is fundamentally the union of our traditional technologies that we had in the original Superdom, our scalable chipset, HPX, IP, VMS, IP, non-stop IP. But you look at it and you say, well, that's also blade system technology. So all of these are come together in one consistent converged infrastructure. So our re-utilizing elements from blade system, from traditional Superdom and marrying them together, now bringing X86 into that mix, marrying that up together, it really is one converged infrastructure. Okay, so you're saying basically that platform is converging and then you have software layer on top of that, which is mission critical as your value at. The follow-up I have on that is I saw, I stopped by your pavilion and I saw a chart that said mission critical today, mission critical in the future, things are changing and you had this little tiny mainframe dot, so they're throwing IBM under the bus, which is good, did I get marketing? And then you had mission critical and then you had X86 outside and then the future, you had X86 moving in and taking up a bigger piece of the pie. There's other things that are going on there, cloud, big data, virtualization, you referenced open source before. How do those pieces factor into the mission critical strategy? So I think that one of the benefits we have with converged infrastructure is that if you look at HP's cloud offering, it's all based on converged infrastructure and the ability to offer that. Now because our mission critical, both our traditional established mission critical, as well as project honesty mission critical with X86 is all part of that converged infrastructure. That means that the same tools are using to deploy using HP's converged infrastructure cloud today for non-mission critical applications, well they'll be able to use that same infrastructure, those same tools, those same procedures, same policies and deploy mission critical on the cloud as well. Now you may not deploy a mission critical environment on a public cloud, but I can certainly see you wanting to deploy the mission critical portions of an application that spans and bursts between private and public and you want to have one tool that'll do that. And you know, if we look today, you can have cloud system matrix operating environment for HP UX, I can have on my cloud map, I can have HP UX and reliable database backend and pull in Linux or pull in Windows on top of Proline, on top of Blades and have all that in one map. Excellent. So it's a Lego block design perfectly for the customer to get the flexibility there. Yeah, what I want to do is, I go data centers all around the world and I can see platforms that have been sitting there running a mission critical application for 10, 15 years. Sometimes they're even still platforms where I designed some of the circuit boards inside and it's great to see those things chugging away. But I also think that those systems are just sitting there and they're doing a great job but they're still using the power supplies from 1997. They're still occupying the footprint that they had which was great in 1997. Not so great today. And part of the benefit of being a part of this converged infrastructure is we have customers who have literally lifted up those images and drop them into converged infrastructure. Certainly the premium, the space in the data centers are the premium and power and cooling is huge. But I want to just change gears real quick and talk about the show here at HP Discover. You guys have a big dream works has been pumping up and endorsing big time up on the keynote with Meg. They're a feature in Superdome too. Customer or they Superdome? Are they using Superdome or those guys DreamWorks? I do not know if they are using it as Superdome. Most of my knowledge of them is around the scientific rendering and that's the massive scale out x86. Okay so talk about some of the performance issues then for someone like DreamWorks. They got a run of production. Where does Superdome fit? Help people understand where Superdome fits into all this kind of maps of tech. So Superdome is really focused in on those largest workloads. So if we think of a global ERP instance it's run some of the largest in the world. When you need a platform with the scale and capacity if you're talking about mobility billing for major mobility carriers. How do you have that back end when you have millions of subscribers constantly generating those events? That platform that can handle that capacity and be running year in year out constantly is really where we're targeting those Superdome platforms. And increasingly what we expect is customers are going to want to understand how can I have all those characteristics but of course bring in x86 economics as well. And nearly that's where we're focusing on. So my final question, I know we've got time running down here is there's two types of approaches as purpose built and we were just at the HBase conference because we were into the whole Duke thing and it's a great solution but we call it the tailored suit. Tailored once, fits like a glove for that use case and you really can't use it, right? So, but really the power of what you guys are offering with mission critical stuff is that flexibility. It's really an operating system kind of constant to systems design. What is the vision, what are you guys doing next to expand on that vision of this mission critical flexibility, high performance but yet flexibility around the diverse architectures that are emerging around. You got flash out there with storage and you got big data with all this promise you got Vertica, now Autonomy. It's stuff above you working changing. How do you guys deal with that, what's next? You know, I think for us it is understanding what fundamental characteristics are driving some of these novel workloads and these are so dynamic. Sometimes it seems like these things are evolving so quickly people don't necessarily have a chance to stop and look at this and say well this could be better if, we could better if. And I think that's where we're getting to now. There's sufficient energy in these novel environments that they're picking up real interest, real use cases and it's just sort of coming up for air, coming up for breath now, people can take one half step back and look at something like a massive Hadoop cluster and say okay this is great but it would be even better if between this node and this node when we replicated data we had a fatter pipe or what would it be like if I subsumed all this data into main memory on a large system and I think that's one of the things that's been intriguing for us as we've talked to our customers about Odyssey, talked about the capabilities of the platform and suddenly they realized well you know I instead of having a cluster of several hundred small nodes connected by an expensive but high performing communication fabric I could possibly collapse all that down to a single node with an HP scalable fabric inside of a super dumb two and say oh I have as many cores I had in a couple racks in one box and instead of being connected by a fabric that measures itself in microseconds or milliseconds it's measuring itself in nanoseconds. I think that's your answer. First of all it's great to have chief technologists on theCUBE so it's fantastic and I think that's a direction that everyone's heading. Just my final question, just a follow up is as a technologist who's designed motherboards we just found out back in the day in 1997 as you think you put the number and someone who's in the real tech of Mission Critical where you need quote big iron like functionality to power a lot of the production scale apps. Based upon your perspective from that view what do you think of big data? Obviously big data is being hyped up all over the place. There is some reality to it. There's a lot of benefits, new ways of discovering and getting insights and all the promise of big data. But big data has one of those things where depending on what you're looking at it's got a different definition. So from your definition what do you think about big data? Well I certainly think that it's interesting to look at the explosion of data and wonder what is the value of individual bytes? I know I certainly care a lot more about the couple hundred bytes that represent my bank balance than I do the gigabytes and gigabytes that represent my music collection back home. And so understanding what is, how do we find the value of information and eventually tie this knowledge and this telemetry and reduce that big data down to eventually transactions because transactions are what actually make people money and understanding where things are and the value of data, knowing what we should be putting in terms of resources and research into these things to find that sweet spot where big data actually can come back to big, big monetary returns. And certainly we've been discovering and talking about the fact that big data needs millisecond response time, you're moving batch concepts into real time analytics. It's going to require some serious horsepower. So I think there's a bright future ahead for you guys and thanks for coming on theCUBE. Really appreciate it. Kirk, chief technologist at the BCS group that's the mission critical group within HP. I'll be right back with our next guest after this break.