 Okay, we're back here live in New York City. I'm John Furrier with Silicon Angle, and this is theCUBE. We are here for a special presentation of theCUBE for HP Moonshot. This is where big announcement from HP that's rocking the industry at point where seminal moment in history around servers and the changing of the guard from old kind of high powered scale up machines to scale out open sources taken over the world driven by mobile, social, and cloud. And we've talked about it all day long and breaking down for you and giving the analysis here from the people on the ground here. But really at the end of the day, we talked about how do you get to this new hyperscale environment and that's what we're going to explore here. Again, I'm John Furrier, I'm joined by my co-host. I'm Dave Vellante of Wikibon.org and those of you who follow this program know that John and I love to talk about the services angle. Services is where the rubber meets the road. It's where customers actually get a lot of the value that they realize for their business. And we're here with Scott Weller, a CUBE alum. Scott's the vice president and general manager of the technology service group within the EG business unit of HP. Scott, welcome back to theCUBE. Yeah, thank you. We've heard a lot about the innovations today around Moonshot and at the end of the day, it's a no-brainer for us at least is to say that this change is happening. Low power footprint issues in this data center. These are real OPEX hard environmental issues. Green really never really made it to the data center. Although it's kind of a marketing slogan. There are real costs, real issues around power and cooling in the facilities, but yet the footprint issues there. And now you have this whole shift towards software defined data center, which is now, it's marketing right now, but still it's the destination where you're seeing software being, as always the key component, but in a whole another way. So we explored that earlier. Talk about your view of Moonshot because you're within HP and you have a huge set of customer bases that are moving there. You already have customers that are doing hyperscale and are changing over from a 10 year lack of serious investment in IT in terms of new things to now cloud mobile and social. So your take on Moonshot today and what you think it represents for the industry and then how customers are reacting in their plans. Yeah, I think you said it well when you said a seminal moment. I think of it in terms of really an inflection in an industry. I mean, how often do we get to sit here and see this sort of thing happen, let alone drive it. So we're very excited about what's announced today and what's going to come in the future. You know, if I think back historically, the workloads that have been built, the applications that have been built over the last 20 years have really been designed for general purpose processors and there's a lot of advantage to that in terms of workload portability and so on but there's a real limit in terms of how much you can optimize that for footprint, for power and for cost and that's what this new generation of servers is really about. It's really changing the game and saying, look, we want to tune these workloads for specific servers and if we do that, we get all of these advantages that you've been hearing about all day. Now, you can imagine that customers wanted to make use of this technology and do it quickly, but it's new. So they need some education, they need some help to plan the workload migrations including really doing a great job of mapping onto the energy footprint that's optimal. So we have services for that. We're being asked to deliver that, so we do that today and then we have implementation service, obviously customers want to get time to value so we do that and then finally we've got a new set of very tailored support services that comprehend this new kind of technology and how it fits into the data center today because obviously customers are not going to rip and replace, they're going to add this new technology. So how do they get a great support experience from that both at a system level and throughout the data center? Dave and I always tend to have these little, small, little private meetings with the Wikibon community around CIOs and when I was in Boston last week and Dave came back on a red eye from California ironically, I was here, he was there, I was like, that's not the way it's supposed to work but for the most part we ended up meeting and spending some time with a close group of our CIO friends. One has a billion dollar operating budget and the other one has a little bit mid-range CIO and we talked about this and one was like, yeah, we're doing a Hadoop Rupert concept so you're seeing some of the new stuff that are working on it and the other one's like, hey, what's this database situation going on with no SQL? I'm just trying to get my data disaster recovery taken care of so you have kind of like, kind of living in the stone age but the reality is that's the way most people are in corporate IT and then you have some of the bigger shops in corporate IT who are doing a lot of effort in doing proof of concept with Hadoop, looking at this new architecture. How do you, how do you get them there? I mean, from here, getting them from point A to point B and then what are customers need to get there and you have to get a delivery but there's this whole nother support angle, whole nother set of services that have to be invented. What are you guys doing there? You know, in some sense, this is really the same as every other new technology adoption curve and there are players who are early adopters and sometimes it's the small entrepreneurs, sometimes it's large companies that do this but in any case, what we find in our customer base is that they're in every part of the spectrum, different functions, different organizations are at different points along the timeline and so really it's about being an expert first and foremost is who do you turn to to help you understand the promise of the technology to begin with and then how does it really play out given the business they're in, the workloads they have, the infrastructure that they've already invested in and so on, their cash profiles, their P&L profiles, business models, all of that. So that's what we do and we do that on a recurring basis. The thing that's different now is this is about workload analysis and understanding how to map those workloads onto a very different kind of technology. Doug Foster, who was on earlier from Purdue is one of the guests you guys had here on your panel, he's this ex-CIO and we had a nice conversation this day so he's got a split personality and academic now doing some high-end research which is kind of like, you know, having a nice little sandbox to play and probably doesn't have to worry about money that much because they got well funded and then he had experience on the corporate IT side as a CIO, it was a little bit different marketplace but one of the things that he talked about is that with Moonshot it's a zero risk situation you can bring Moonshot in and play with it and get a feel for it if it works, you double down on it. We tend to agree with it because it's a pretty lightweight to work with and it hits a chassis and put 1,800 servers on it so you spend a little dough, a little POC. How does someone do that with HP right now? So that's a nice selling item for HP. Okay, I wanna take that chance. I'm willing to bring HP and the cost savings are ridiculous. So that's a check box. HP's a big name, brand name, they're not going anywhere. They're a good supplier. Okay, come on in. What do I do? What do you do when you walk into that engagement? I mean, I'll see if there's some money involved. It's not a lot of untold. Well, I mean, it starts with a great facility that we have on the Houston campus. It's the Discovery Lab for Moonshot. It's a fabulous environment. Customers can come there and learn about it, actually play with the technology. But in situ, in the customer's environment, this is where, yes, you wanna do a proof of concept or invest in some of your rack space to understand what you can do with the Moonshot technology. And really, I think it is a small investment for what could be a game changer. We believe as a game changer for the entire industry, but specifically, we know that there are configurations of Moonshot that are going to be a major improvement in every way for specific verticals and vertical applications. We've talked about some of those to date. There are more to come, but in any case, it's the traditional advising around how to migrate workloads, how do you get the best energy footprint, how do you implement the thing so that you get time to value, and then how do you have a great support experience that's more personal, proactive, and simplified, which is our core support message. So they come into the Discovery Lab, not to be confused with Purdue's Discovery Lab. Correct. That's kind of the same thing. So they kinda come in, they can kinda play with it, kick the tires. But you have specific use cases that you have with them that you can say, hey, you know, given your footprint or your configuration, here's how we can get it in there and integrate it into their environment pretty quickly. Have you seen some of those use cases, and can you share some? So the product that we're announcing today is really tuned for sort of the edge of the web, if you will. And, and... You mean devices and like... No, I'm talking about the web front-end to, for example, a web retailer. Oh, okay, got it. So that's what we're talking about today, but we do have a roadmap and a plan for different configurations that are tuned for things like big data analytics and so on. And we've talked about some of those today. So you can begin to explore those different kinds of use cases in that environment, but again, the best place to do that is within the context of a customer's business environment so that you get some visibility and some insight into what the real value can be. So to me, this, and we've been talking about this all day, is this architecture is gonna, fundamentally take over the server, whether it comes from HP or another set of competitors, which I would fully expect you're gonna get. And we talk a lot about hyperscale for obvious reasons. A lot of little smaller processors, right? So you think hyperscale. When you look at the hyperscale guys got there from a management of technology perspective, they're just trying to drive unbelievable levels of automation down to the point of if something breaks, they just shrug, they put it to all the guys and then throw it out the rack, throw it into the wood chip or as they say. Do you see your traditional enterprise customers trying to sort of simulate that to some level and to what degree and how does that change the mindset of how you go to market with services? Yeah, I think it's interesting. There is maybe a dichotomy in perception of what's going on around hyperscale versus the reality. So we all know about the big name tier one hyperscale customers, the brand names on the web. And then the assumption is, well, that anybody who's interested in that proposition, that value is kind of on their way directly there. Don't stop, just get me there. But the fact is that the real business is out there beyond those players have an enormous portfolio of assets, all kinds of applications that run different ways. And so what they need is they need a migration strategy. We've already seen that in virtualization. We've seen that in private cloud. And this new technology is going to be a similar path for them to understand, okay, this is a net new technology. It asks me to think very differently about the workloads and the assets that I have and how do I get the value out of a proposition, this kind of technology where it says, I'm gonna tune a server to the workload versus the other way around. How do I get the maximum amount of that? Which workloads are even candidates for doing that right now? So that's really the transition we're going through. And yes, we do expect to sell a lot of moonshots to the big brands on the web. No question about it. But everyone else is looking at this technology as a way to get the value to see some of those performance and cost and energy benefits, but it's not gonna be a wholesale move. It's going to be a thoughtful approach. Are you suggesting that the big brands that you sell moonshot to have the resources, many of which you would bring to the customers that don't have those resources? Is that a fair characterization? In other words, you've got a lot of folks in the traditional enterprise that don't have 1,000 PhDs running around. That's sort of what HP does for them. Is that a right characterization? Well, so if you go back to the question about automation, I mean, the very large scale out customers definitely are automating in ways that don't happen in a lot of the rest of the industry. And yes, they have different service needs. It's not to say they don't buy service from us. They need different kinds of service. They look at different ways to purchase OpEx versus CapEx. They look at having an onsite kind of sparse facility that other customers may not wanna have and so on. They expect to have, when they do have an issue and they call us, they expect us to have a dedicated call center for them. So we do all of those things. But again, those are just different kinds of services for that market, not no services. Yeah, how much of your business that leads me to another question I'm just envisioning what you just said about the spares and how much of your business can you repeat versus how much of your business is, hey Scott, we need this that you've never heard before and you say, yeah, I can do that. Well look, I mean, every big company in the industry can write a custom service arrangement. And sometimes we do that, but over the last several years what we've done has been very diligent about reducing the number of those because after all, we wanna create a very consistent and great customer experience in the ownership phase of Moonshot technology or any other technology they buy from HP. And so to do that, you have to have a consistent set of processes, a consistent set of offers in your portfolio and you have to do that on a global basis if you wanna be a company like HP. So customization isn't your friend if in those cases, although we do it when the customer needs us to do it. But in any case, we have really kind of pared down what we do, we focus what we do to get more of a building block approach rather than a raw customer approach. A lot of what you were describing before in terms of thinking through the energy strategy is more consultative in nature. It is. Primarily support services. Where do you guys draw the line? I mean, you specifically in your group and then what kind of new consulting services are you seeing emerge as a result of Moonshot? So just thinking holistically about technology services on the consulting side, that team is doing what it traditionally does which is advise about how to get the most out of new technologies. How to adopt them, how to map it onto their business and then how to get it deployed and then implemented. So that's what the consultants do. And around Moonshot, we've developed a whole set of services specifically for Moonshot to do just that. On the support side, again, we've tailored our support services so they comprehend Moonshot. I mentioned a dedicated call center. If a Moonshot customer has a question or issue, they need to have a backend call center that knows about this technology in a deep way. And since it's so net new, we had to dedicate those resources. And I can give you a lot of other examples, but the point is we've stood up our value chain to stand behind the Moonshot product. We think it's absolutely going to change the industry and we're ready for that. So I also wanted to ask you about the following. So customers that we've talked to, John, you said the same thing when you first saw Moonshot. You said, I want one. The customers do. It seems really awesome product. When customers see this thing, they say, oh, I know exactly what I want to do with it. I want for this set of workloads, I just give me some because I got to have that. Not necessarily thinking through the implementation nuance associated with that. So what kinds of things do you anticipate or from your experiences over the last, they say 15 months and then so the early customers, maybe even your own internal experiences. What are you seeing there in terms of what's different as far as the implementation and the uptake and things that you're advising customers? Yeah, so I think, again, it starts with a bit of a new notion. I mean, we've had this long history of developing workloads against general purpose processors. And so first of all, this is a shift, right? It's a shift around that mindset of how do I let the workload dictate the kind of server I'm going to build and ask HP to build for me. So that's the first step. Then it's the question of what workloads are really amenable to this kind of environment. And again, back to the adoption curve and the transition in the industry. Yeah, I think over time, this technology, this model is going to become pervasive. But at the current time, there's so many workloads that are running massive critical processes and companies that probably, for one reason or another, are not amenable at this time. So it's a question of how do you segment the workloads that you have? Which ones do you want to try on this new technology? And that's what we advise about. So we saw this wave of efficiency in the last, say five, seven years, eight years or whatever, through virtualization. Right. But what I'm sensing is that by allowing finer granularity at the server level and allowing this tuning the infrastructure for the workload versus the reverse, you're putting forth a vision of taking that level, that efficiency level, to new heights and driving new business value. Are we understanding that correctly? You are. And this is sort of the next step beyond virtualization? You are. I mean, in a sense, virtualization was a way. There were a lot of aspects of virtualization. And this is a narrow description. But it was a way to get more performance out of the investment you'd already made. Isn't that a utilization plan? Absolutely. In fact, we've all known about the numbers of utilization prior to virtualization, right? Quite low. And so yeah, I mean, in that narrow sort of lens, I would say yes. It was about getting more value out of the asset. And in a way, this is the next step that says, the mainstream workloads have not been written this way. There have been all these niche industries, avionics, and so on that have written to targeted processors. This is a way of bringing that into the mainstream and saying, look, given where we are in the technology curve and in history, we've got this new set of technologies and processors and so on, where we're ready now to apply the same kind of concept in a general way. My final question, as we're getting close on time here, Scott, is a question on the webcast. I think it was to Savas. How has this changed your services that you offer? I want to ask you the same question. How has Moonshot, as you're announcing a new product, but that notwithstanding, how does Moonshot change your set of services within the support and consulting organizations? Because it's a game changer. It is an inflection point. It is a seminal moment. So how does that change or enhance what you are currently doing with your customers within HP technology services? So what I want to do in the support business is to really to continue on this path of simplification for customers. In some sense, if you have different servers for different workloads, customers may think of that as bringing in an additional level of management for them, a different level of ecosystem for them. And so what we're on the path to do is to really make all of that sort of a seamless experience for them. This shouldn't be about bringing more complexity into the environment. This should be about bringing all of the value around energy consumption, around cost and space efficiency, and do all of that without adding layers of complexity. So part of what we intend to do, and we've already begun to do for our customers who are early adopters of this technology, is to really keep it simple, keep the experience very proactive for them, and personalize for them, and then really just to remove all of the complexity that we can out of that environment. It's a great opportunity. Certainly, we heard from one of our guests early, it's going to be extra innings. When I asked, what inning are we in, they said, I don't know what inning we're in, but whatever it'll be, it will be extra innings. Meaning, there's a lot of growth there, and a lot of transformation. So Scott, thanks for coming back on theCUBE alone, breaking it down for us. Obviously, it's going to have a huge impact too. Channel partners, support services, a whole new set of portfolio. I'm sure this will just create a massive growth organically of new services, helping customers get to a new, modern infrastructure. This is theCUBE, a special broadcast live from New York City. John Furrier with Dave Vellante. We'll be right back with our next guest. So stay with us here at New York City. Right back.