 From London, England, it's the Cube. Covering Discover 2016 London, brought to you by Hewlett Packard Enterprise. Now, here's your host, Dave Vellante and Paul Gillin. We're back at the London Docks. This is the Cube. We're here live at Excel London HPE Discover 2016. My co-host, Paul Gillin, is back after a nice short break and Alan Andreoli is here as a Senior Vice President and General Manager of the Data Center Infrastructure Group at HPE, new role for you, congratulations. Great to see you again. Thank you, Dave. Great to see you. Yes, so your domain is expanding. You're now an expert in networking and storage and of course servers and you're number one in servers and now you need to do that again in networking and storage, right? No problem. That's the objective. And as everything is converging, we want to continue to be the leader and the south leader in the data center overall. So we are making great progress. Well, I mean, in servers you've done very, very well. People have been trying to knock you out of that number one spot for a while. You know, there's chatter about units, but the real measurement is revenue. You've been number one at revenue for a long time. That is the primary KPI. Is it not? And listen, I got some good news this morning actually. We, Gartner, just published their results for the calendar quarter and we are distant number one in revenue, in compute and we also continue to be number one in units. So that's great, right? Despite the ramp up of the ODMs with the cloud and so on, we continue to be leading the market by a long shot. So that's a very good thing. Although this is not our objective per se of being number one for the sake of being number one, we are in the business of doing business and be profitable to be able to generate investments so that our customers get solutions that match their needs for the long term and to invest on innovation. So we're doing that, but we remain the number one on the market and that's good. You said that's a byproduct of your effort with customers in your- I think it's been for 80 quarters. It's something like this. Yeah, yeah. So that's great. That's good. So talk about the SGI acquisition. Relatively small acquisition, but strategic for- Very, it's exactly the continuation of what we've been discussing before, which is our segmentation strategy. So we've been defining the poachers of the market that continue to grow and we are doubling down. And so we've been growing last quarter 45% in HPC. We've been growing 17% on mission critical applications based on x86 driven by SAP HANA. And so now with SGI, we becoming the expert in these areas. Not only are we the leader in volume and in size, we're also becoming the niche expert on the very top specialized high end of the market in supercomputing. Is it a niche or do you see a broader market opportunity here where enterprises, this technology actually becoming part of the mainstream enterprise- And a big data play. So it's a great question. So it's both. So it used to be a niche and this niche has got to be taken mainstream and that's what we can do as a large scale vendor because the craze and there's the eyes where niche players who are playing in a small sandbox and we can play in this sandbox with the scale of all the technologies that we have and expand this commercially. What we couldn't do before, for instance, is to win very, I would say, secret federal programs. And now we can. So we can play in very, very unique and advanced technologies and decide what we want to take to a mass market. In the era of big data, HPC mission critical will become mainstream. In the era of deciding what stays on-prem and what goes off-prem, what will stay on-prem will have to be mission critical and high performance primarily. So we see this very strategically. Yes, it's a small acquisition Dave, but in reality it will have a big impact on how we become more and more an expert of what is specialized and what is mission critical. And I was encouraged to hear Meg talk about it. I forget the exact number, Ellen, but she said you'll compete now for some number of the top hundred supercomputer performance. So you're competing against China Inc, in my view. China is trying to become self-sufficient. They've got, I don't know, four out of the top 10. Is that important? So let me give two responses. First, made in the USA. Everything SGI done is made in Shippewa Falls, Wisconsin, USA, and there is a great gang of people there. The metal is made in Wisconsin or Minnesota. They don't go to China. They made everything locally. So it may sound like interesting given the times. So this is technology for the country. This is technology that is clearly differentiating us from the rest of the world. Then you look at China. So if you exclude China in the top 500, we have 41% market share in the top 500. If you include China, we are still number one with 140 systems out of 500. So we are now the distant number one in this market that we decided to attack as you remember strategically three years ago. Then the question is, okay, it looks like there is a big number suddenly and growing number of Chinese supercomputers. First, they are captive. No western companies can access. And second, are these benchmarks to satisfy some political objective of being the world leader of whatever? Or are these real supercomputers doing real world? I don't know because we don't have access. We don't know what they do with them. But we know from the ones we know and we know what they do, we have 41% market share. And this is much higher than we've ever had. Yeah, and regardless, the strategic implications are, I think, significant. Curious about the trend toward multi-GPU systems that are suddenly so popular in scalable and super QE type applications. Is Silicon Graphics a play for you in that area? Or are you playing that work? Yes, of course, they play with GPUs and we are expanding our relationship with Nvidia. And actually, in my segmentation work, I have now expanded HPC to be HPC and AI. And GPUs basically enable machine learning and artificial intelligence. And so we, Bill Mannell, whom you may have interviewed in this event, will carry this flag for the company all together. Is that how you're positioning, how you will position the SGI going into the enterprise? This is essentially machine learning and artificial intelligence machine? So if we talk products, there are two key values for us of the SGI acquisition. Beyond the fact that it's made in America, beyond the fact that they have hundreds of federal certified people and so on, they have two flagship product lines. One is called ICE, which is their fifth generation of supercomputer. And the other one is called UV, which is their first generation of mission-critical in-memory systems. These two key products will be carried over in our product line moving forward. So you remember when we invented our high-disconnect technology, which, you know, this is great and we've been very successful with it. But SGI has a technology which is also very, very interesting and very cost-effective and more pervasive in the industry because they're a fifth generation. So we're blending the two teams together. We're keeping engineering, we're keeping manufacturing, we're keeping the development sites and manufacturing together in the USA, and we're going to take them to the next level, enabling them to go to commercial spaces where they didn't have the scale to go before. Will SGI be part of the composable infrastructure? Yes, it depends how you call composable infrastructure. SGI is actually very specialized hardcore infrastructure. So composable infrastructure is more how you adapt, configure fluid pools of resources to match with your cloud enablements. SGI is more on the specialized side of the business. There are three groups, if you want, moving forward of infrastructure. You have the generic infrastructure, the classic 886, which we've had for 25 years, the Proliant and so on. Then you have the specialized, which is doing something sort of a kind of a very critical mission and unique mission. This is where SGI actually plays. And then you have composable, which is fluidity and that adjustment of pools of resources for the cloud. So I would say it's slightly different, but SGI will be integrated in the overall continuum of our product offering. Let's talk about storage a little bit. Part of your purview now. And historically, let's say the last 15 years, server-led companies have not been number one in storage. It's been EMC and NetApp. It looks like the market's changing a little bit, but historically, you know, HPE, HP, IBM, Sun, going back to digital, have never really been able to crack that number one. Are things changing where a server-led company and storage is coming closer to the server and flash is bringing it closer. Memory-driven systems are bringing it closer. Is that pendulum finally swinging where a server-dominant company can be a dominant storage player? Dominant, number one. So the boundaries between storage and compute are definitely blurring. So you have external storage and you have internal storage. The internal storage is exploding. So in our next generation of compute, our Gen 10, the space that is dedicated to memory or storage options is much bigger. The number of buses that address memory or storage are higher. So compute and storage in racks are becoming completely very. And then external storage continues, basically, but becomes more and more flash-centric. So we are just passing the 10% flash line right now. We've been growing 100% year on year with our three-par, all-flash storage. That 10% is a market number, right? Yeah, the market number. Three-par is 50%, right? We are growing 100%. But the percentage of three-par that's all-flash now is 50%, right? It's about that, right. And so we are growing 100% on, what is only 10% of the market. So you can see how much potential for growth we have. And we have been growing very, very strongly with our compute-centric storage, internal storage as well, which has made us number one when you cumulate the two. Right? Yeah, when you cumulate internal storage and external storage. Even with the Dell EMC merger. Yes. You were number one. We were before the merger. I don't know what's the latest when you add the two of them together right now. So it's a very strong growth we have. Our strategy in storage is continue with the growth of three-par-flash. And this has been, I mean, three-par is now really a super franchise on the market. The gift that keeps on giving, I say. And we are now offering it as a flex capacity. I don't know if you had Bill Filby coming here, but you know, three cents, a gigabyte per month. I mean, we are now offering the flexibility of the public cloud completely on-prem with three-par-flash. So this is going very, very well. Then we are reattacking the entry storage with a new family of MSA and a new family of entry storage altogether. So that's, we think is going to give us another big boost. And then we're going object storage, which is internal storage, but I moved the internal storage offering which was part of the Apollo line into the storage group. So now we have total storage as one integrated group. And our partnership with Cality is very successful in taking files into object format and therefore expanding capabilities for large-scale, hyper-scale archiving and storage. Which you need to mimic the public cloud on-prem, right? So we are in a strong position in storage and we are a little bit more ambitious. So we have strategies that go beyond organic. We are looking at the next horizon. So storage is an area where we are very confident we're going to continue to be in a very strong position. Are you suggesting that there's some potential M&A in the future? For the company? As a whole? Yes. Well, you have to ask Meg, but we are... What we did over the last year and a half has been refocusing the company on the data center and on the edge. And so now we are very, very clear on where we are. From that point, we have generated a lot of cash and Meg has done a phenomenal job in creating a very, very healthy balance sheet which allows us to be acquisitive if we want to. 2.1 billion in cash with 3 billion back to shareholders last year. And she said yesterday that by the middle of next year we probably will have 11 to 12 billion dollars in the balance sheet. Yeah, yeah. So very different of what some of other competitors are doing. So it's a long-term debt for the first time in a while. So we've talked about this a lot. It's a lot of money. The 3-par storage as a service, essentially on-prem of storage as a service, is this the beginning of a trend where we see other services, other products out of HPE on the compute side that are offered on a similar basis? I'm sorry, I didn't hear your question very well. The new storage as a service, 3-par, 3 cents per gigabyte on-prem. Is this the beginning of a new strategy that we'll see in the compute side as well? Totally. So all of our infrastructure, we want to give customers the best choice, the right mix for their hybrid IT journey. So if they want to go on the public cloud for some workloads, we will help them doing that with our synergy layer. If they want to be on-prem, we will give them the same flexibility and the same cost benefits, which includes flex capacity. So you can either say anything I buy from you will be at X cents per gigabyte or whatever, or you can say I'm going to buy half of my capacity, but then I want to be able to have big demand for 50% additional, and I want to have, you know, and that's what we're going to offer. How are you incenting your sales force to sell this? These as-a-service options are not, it's a change of thinking for salespeople to sell that. What are you doing in that area? Well, they will be paid, it will be transparent, it will be neutralized for them, and whether they sell as-a-service or they sell with the same mobile. Yeah, I would think you make more money over time as a service. It might not show up in the income statement as attractive in the near term. But it's not those big lump sum payments. Right, it might not be as attractive near term, but long term, I would think it's very attractive, business model. Yeah, and it becomes an annuity business, right? Which is good for both parties. Definitely. But this choice has got to come from the customer. Do they prefer, I mean, do you want to rent or do you want to buy your house? You know, it's like capex, OPEX, it's a choice you make with your car. Do you want to lease your car? But we will offer all these options and we really want to be the partner, neutral for our customers to go to these hybrid IT journey. Yeah, and you've got the financial resources to deliver that capability. We have HP financial services. I want to talk about, we're out of time, but I want to talk about networking, give you an opportunity to talk about that business. We just had Dominic Wilde on with Arista and a customer. It seems like you've got now the three-com piece, the Aruba piece, and now the partnership with Arista, the portfolio is pretty broad. Talk about your comfort level. We have way to go. We have way to go. Yeah, okay, so you got a very entrenched competitor with two-thirds market share. What's the strategy? So, first let's look at the partnership with Arista. Cisco has lost more than 10 points of market share in the data center over the last few years. Arista. In networking or in the search? Data center networking. During the same plan frame, Arista has moved to be 10% of the data center networking business. So, they are the winner. When you have that scale and they're getting to be around a billion dollar in size, and you are in that type of situation, you have different ways to scale. You keep investing like crazy in your own. Go to market and so on. Or you make a partnership that allows you to get the reach that you don't have by yourself. And that's why it's a win-win. On our side, we have room to get as big in networking as we are in computer and storage. On their side, they need to scale up. So, all of our channels are now going to be activated on Arista. And this is going to be the next jump for Arista. Their strategy of being software defined is exactly ours for storage and for compute. So, it exactly matches our strategy. So, this is a partnership that is defining for both parties and that will allow us to take our networking presence end-to-end, together with the rest of our infrastructure to the next level, and allow Arista to move to the next 10 points of growth in the data center. A lot of upside there. A lot of upside. Alan, we have to leave it there. Thanks very much. Love having you on. Great guest. Thank you. Thank you for all of your business in the marketplace. Thank you. You're welcome. All right, keep it right there, buddy. We'll be back with our next guest. This is theCUBE. We're live from Excel London, HPE Discover 2016. We'll be right back.