 Hi everybody, we're back live in Las Vegas. This is Dave Vellante from wikibond.org. We are covering a number of events today, the HP Gen 8 announcement. We got our man Jeff Kelly's at the Data Warehouse Institute. Stu Miniman's been hanging out with some of the guys from the VMware Partner Conference. A lot of action in Vegas this week and we're here with Richard Foshera, who is Vice President and Senior Research Analyst at Forester. Richard spends a lot of his time looking at converged infrastructure and data center issues, but also Richard spent a lot of time with customers as well. Welcome to theCUBE. Thank you, this is always a treat. Yeah, good to have you. Well, let's see, let's just get right into it. I mean, Gen 8, HP's talking about the importance of automation in your assessment. Have they hit the mark? Yeah, this one is pretty spectacular. I mean, when I first started hearing about Gen 8, I was expecting the usual, we got the next generation Intel processors, we got, you know, maybe we've made some enhancements to the system more, yeah, faster, probably more power efficient, maybe we've got some better hardware management and fast stuff. But this is really, this is like a, this is like a wide frontal assault on all the things that I've been hearing for a decade as significant cost and operational pressures on the data center operators. So do you have this force to have a scenario around so-called next generation servers and does this fit into that mold? We don't have a formal scenario document, but yeah, this definitely fits into, you know, fits into what an intelligent vendor would be doing. Let's come up with one right here, real time. Well, the scenario is, you know, I think they approached it the right way and I have to sort of issue the disclosure. I was, for four years, I was director of blade system strategy at HP, so I know this from the inside. Yeah, okay. You know, you got to listen to your customers and what I'm seeing today are things that we were hearing, you know, when I was there from 2006 to 2010. And as an analyst, I've been hearing for a decade from customers of every vendor. So the pressures on cost operation, particularly OPEX, as CAPEX has become less of an issue, the pressures on OPEX and power efficiency have become extreme. So what they've done is they've answered, they've answered the requirements their customers are putting on them and that's kind of the core of what a smart long-term server strategy is you better answer your customers' problems or you're not going to do so well. And what they have done is applied some very appropriate technology solutions and are wrapped it up with some services that's going to be very hard for other people to duplicate. Yeah, so they're really following the money on this one to your point. It's not in the CAPEX, it's in the OPEX and that's really where this is aimed. But you made a point in the panel that I thought was right on and I want to have you talk about it a little bit which is the messaging that we're hearing is there's a lot in there on unplanned downtime and sure, a catastrophic event is going to be expensive but your point was that planned downtime is actually on an ongoing basis a lot more expensive. That's where most people are spending their money. Can you talk about that a little bit? Yeah, it's not as visible because it's not catastrophic. It's built into the budget, it's built into the schedule and the way you build it in is by having more resources and spending your precious OPEX dollars on managing downtime and maintenance and configuration changes in a non-disruptive way. And that's not cheap and it's not easy. So it's a hidden problem because it doesn't result in the company's sales portal going down or calls at two in the morning. But it's a huge budget item and a big operational issue and it's those things in combination that suck away the resources from being able, for example, to implement a new project for a brand new business requirement. So Richard, what are your clients telling you in terms of how they want to drive change in the data center? Well, most of my clients would be very happy to be in a position to actually drive change instead of reacting to external pressures just to be in the position to drive the change on a planned proactive way as a step forward for a lot of organizations. They typically want to try, in their hearts, they want to try and stay ahead of the business requirements, but usually it's like the old thing about when you're up to your ass and alligators, you hard remember you came there to drain the swamp. So the other thing I want to talk about is the energy piece. Obviously there's a lot of the thrust of this announcement is aimed at the energy. Having said that, a lot of CIOs, and I think actually, I think Forrester had some data a couple of years ago that I saw that about, and I know the number's been ticking up so maybe you can update it, but I think at the time maybe 15% of the CIOs out there actually saw the power budget. Is that changed? I have not seen a good hard number on it. I think the number has gone up because I hear more of them are more aware of it. I'd be comfortable in saying that even if they don't get the bill in their budget, they're much more aware of it than it used to be, but there's still some structural problems in a lot of organizations. There's still a lot of companies where the person that pays the power bill doesn't have a direct connection to the IT operations. The facilities has always been sort of a separate walled off domain. So it's probably fair to say the situation's getting better, but it's certainly not perfect yet. So as a person who follows data centers quite closely, you'll appreciate this quote. A CIO said to me the other day, last month I was in New York at an event, and he said to me, my budget's not growing at 45% a year, but my data is, and I think it underscores the situation in a lot of data centers. On the other hand, you've got these cloud service providers that are growing like crazy and their budgets are growing like crazy. What do you see as the juxtaposition between the cloud service provider and the traditional data center? Is the traditional data center manager under pressure to perform like a cloud service provider? Absolutely, and the good news, bad news, good news is, bad news is he is under that pressure to perform like a cloud provider. The good news is that I think there's going to be definitely sort of a merging of the DNA over the next several years as companies like HP bring to the enterprise IT people some of the capabilities that the cloud providers have in terms of some of the economics, but absolutely the yardstick against which they're compared now is more and more the cloud provider, but I think a lot of that technology is going to be available to them. The ability to rapidly deploy servers, for example, that's a thing that a cloud provider has to do really well, but now you see with the Gen 8, there are opportunities for the enterprise IT people to get very efficient deployment of their servers. The operational monitoring and so on, those are becoming more and more accessible, but the pressure will remain and in terms of overall demand for servers, which is a little off your topic, but whether the application lives in a cloud or in your data center, it's still got to live somewhere and execute, so cloud is not going to reduce the demand for servers. If anything, it's going to increase it. Yeah, I know, that's an interesting topic and one that I used to be a number cruncher at IDC and I always noticed when prices dropped, demand went up, it was an elastic market and I would think of anything, it's going to increase it, but I want to follow up on the cloud service question is, what do you tell your customers or what are you seeing in your client base around chargebacks because they have to act like these cloud service providers, but only maybe 20% of them at most do chargebacks. Do they have to start doing that? The answer in the classical consultant fashion is it depends. It depends in an ideal world, we like to say, the definition of cloud includes elasticity, always on fine-grained chargeback. Now, my colleague, James Staten, and I go around and around on this in various fashions, but I don't think that perfect chargeback is always a requirement in a corporation as long as the economics of the solution are reflected somewhere realistically in the budget, it doesn't have to be a perfect chargeback. It can be an allocation, it can be a fixed fee, as long as in the end it's a realistic reflection of the resources. It doesn't have to be a perfect per CPU-minute chargeback. However, you're right, most people don't do it, it's a challenge and designing the right kind of chargeback is a challenge, but I'm not, if we were a lawyer, we would be saying strict constructionalist. You know, I'm not a strict constructionalist on chargeback. I think as long as it makes business sense and it makes the resource consumption visible to the financial management system, it doesn't have to be a perfect chargeback system. Yeah, so a cloud service provider has to obviously run IT like a business because their IT is a business, you're saying an internal data center manager has to be business-like and very efficient like a cloud service provider, but doesn't necessarily have to have all the metrics in place and all the systems in place to thrive. Exactly, because remember the chargeback itself and the process of administering it cost money. And this has always been, historically, there's always been this balance between fine-grain metrics and what it costs to implement the fine-grain metrics. You know, and this has been a juggling act since practically since the first day I got a paycheck in the computer industry, that fundamental problem hasn't changed. If you make the metrics sufficiently granular and sufficiently granular and fine intervals and a very tight financial control system, you can end up spending more money on the fine-grain metrics than they will save you in internal costs. Yeah, and you could end up getting shot by people in your organization, too. All right, my last question is on a couple of dimensions, on a scale of one to 10, how relevant do you see this announcement? Oh, I'd say about 9.321. 9.321, so pretty relevant. Yes, a very, seriously, a scale of one to 10, you know, nine, because nobody wants to give it a 10 and anything less is probably less than it should get. This is a very significant announcement. It's a big deal because of the breadth of integrates, the breadth, the number of places they've touched the fundamental problem of data center operating expenses is what really impresses me. Each point technology in its own is interesting. There are cases, you know, other people have some technologies and products that address some of these points, but there's nothing I've seen that addresses the whole breadth of them the way they did, and I really think they did do this from a very user-centric perspective. So yeah, the scientific analysis in the end is this is a big deal. Good. All right, Richard, well thanks very much for coming inside theCUBE and sharing your perspectives. You bring a great independent view. Forrester continues to do a good job, so thank you very much for the great work that you're doing in the community, and would love to have you back sometime. Well, thank you. It's an honor and the analyst fraternity to be in theCUBE. Oh, well, as I say, you're welcome back anytime. Thank you. All right, take care. All right, everybody, we are going to be right back after a quick reset. This is Dave Vellante. We're live from Las Vegas, HP's Gen 8 announcement, and keep it right.