 Okay, we're back live here in Las Vegas for IBM's Information on Demand 2012, their premier conference around information management, high performance scale systems. Of course, big data, big data's hot. This is Silicon Angles theCUBE, our flagship telecast. We go out to the events. Do you want to see them from the noise? 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 Peter McCaffrey, who is director of marketing for pure systems. IBM's new line of systems as of last April in expert integrated systems, pure systems, pure data, a recent announcement. IBM's answer to the converged infrastructure play. Peter, welcome to theCUBE. Pleasure to be here. Good to have you here. So we're hearing a lot about expert integrated systems, pure, you guys had an announcement a couple of weeks ago. We did. All around pure data. Yep. So you've really, you were a little bit late to that integrated systems party, but you came in with a very strong offering and with all guns a blaring. We did. We actually brought out the family in April this year. Now it's interesting, we've been late to the party. You could argue we've been doing integrated systems for 20, 30 years all the way back to the AS 400, but this was a little different. And when we announced the family back in April, there was a specific problem that we're setting out to solve and it's one that many of our data center clients face and if you think about it, a lot of CIOs today are in the manufacturing business. They run assembly lines of IT components where they're putting together the hardware and the software, takes time, takes skill, takes resource, never ending. And we said, well, we do that too. Why not put it all together and take a lot of that burden off of our clients? Now, I said we've been doing it for a number of years. We have, what we tried to do different here is for us, it started with the integration. If you look at a lot of the other offerings out there today, they start and stop with integration. That was necessary, it was an important fundamental for us, but we wanted to build on it. And we built on it in two ways. One is, we said, we really need to do a better job of capturing best practices, expertise, and find a way to build and automate that right into the system. So this notion of built-in expertise became an important characteristic of pure system family. And the second was, a lot of the offerings out there today, it's about speed of deployment. And that's nice, but once you're deployed, we'd like to continue to deliver value. So for us, it was also about how do we simplify each stage of the IT project lifecycle from upfront design and procurement, to ongoing management, to backend upgrade and support. And so we announced our first two systems, Pure Flex and Pure Application in April. And then, of course, a couple of weeks ago in Boston and Singapore, we announced three models, we announced pure data with three workload optimized models for data. So when you thought of doing expert integrated systems, I'm guessing it wasn't after some of your competitors announced theirs, you've been working on this for a while, presumably. But what was the conversation like inside of IBM? Because if you're going to embed expert knowledge into expert systems or integrated systems, it potentially means you're going to sell less services. Did that create attention? And how did that all get rationalized? It was, hey, we're going to just do a better job that's going to raise the tide for all boats? Was that the philosophy or talking about that a little bit? The idea there is the service mix will change over time. So today, where a client may be investing in our services around integration and setup, instead, those dollars could be used and more focused around application layer services or higher level services, maybe around an industry application. So it's allowing them to kind of get past the plumbing and onto the higher value thing. So it's really, it's a win-win in our services organizations. Well, it's not unusual for them to see the mix of services change with time. And it's kind of interesting, our ecosystem partners feel the same way, whether there are independent software vendors or some of our traditional resellers or distributors, they're in the process of transitioning their business models as well. And so in many cases, they're moving to higher value services. In some cases, they're moving their models to a cloud-based model. And they're seeing a lot of value in these types of systems, both in terms of speeding the deployment to their clients set, but also using it as an accelerator or a building block to evolve their business models. And we've got some great examples of managed service providers who are leveraging and purchased our PureSystems technology and are leveraging it to get their services out in a cloud-based model. So PureSystems is pretty robust. It's more than pretty robust. It's probably the most advanced or at least broadest portfolio of integrated systems in the industry that people could debate that. But so let's lay that out for people. So essentially you have the spectrum of single block of infrastructure that's a single SKU, you drop it in, any color you want as long as it's black. And then you've got the other end of the spectrum, which you see in the marketplace, a lot of reference architectures that can do any hypervisor and any server, et cetera. Now you probably don't have the any server version, but actually you probably do. We do, we actually do, you're right. We offer a sort of a blended approach. So you have pretty much the entire spectrum and you allow other people storage to attach in, if I understand it correctly. And you're the only company that can claim to have these expert patterns embedded. So I'm trying to sort of paint a picture of that portfolio and really understand those expert patterns, that expert knowledge, how much of that is real and how much of that is just really great marketing? Well, let me start with the concept of the pattern. Because you're right, it really is what separates us from the rest of the industry in many cases. Very simply, a pattern, you can think of it as a best practice or almost a recipe for success. Their patterns will come at different layers. You can have patterns around the infrastructure, the middleware or the application. I find that the application sometimes is the best way to provide an example. So if you have one of our partners in for, they have a robust ERP based solution, they're doing application patterns. But if you look at what it takes to deploy an ERP system today, there's quite an application topology. There are app servers, there's database servers, you have to be at certain levels of the middleware. You have to have certain, there's certain capacity requirements to run that configuration at optimal performance. So you're right, a lot of people today will go out and they'll build a blueprint or they'll build a reference architecture. What we've done is provided a way to capture that reference architecture and codify it into a pattern that is then read by the machine. So pure systems can, you know, a client can basically go out to a catalog, select an application pattern, and it's really pretty much a push button approach. They'll specify a few policy parameters. Do I need high availability? What are the performance boundaries of my ELA? So they'll set up a couple of, they'll answer a few policy-based questions. And then based on the answer, we'll go ahead and deploy that application based on that best practice and the system makes decisions relative to what kind of resources it needs to apply to that configuration. And so all of that is really taken out of the hands of the client. And what they're left with at the end is a best practice deployment without all the risk. Now that's the deployment process. I mentioned earlier that the value doesn't stop with deployment. Once that pattern is deployed, the management software continues to monitor performance of the application. And so if it falls outside a performance boundary, it will make decisions to spin up, let's say, additional app servers to get past that spike in demand. So we're constantly monitoring for versus policy and we'll make some decisions on behalf to address the spikes and valleys in demand. And then, and this is where I find a lot of customers get excited, is a lot of times what stands in the way of upgrading to that next release of the software is the risk and complexity associated with upgrading that application topology. You have to know what fix packs are running and what version of those fix packs, what levels of the middleware. Our clients spend a lot of time researching that with the vendors and frankly it takes a lot of time, a lot of effort and there's a lot of risk associated with it because if you make a mistake you could knock out that application. So with patterns, upgrades are now done at the pattern level. So a vendor will provide an upgrade for the entire application topology all at once. So instead of dealing with literally hundreds of separate points of maintenance, you have one. And that significantly streamlines that back end process. And that's a compelling value proposition. It really attacks the IT labor problem that you were talking about before, particularly the skill set side. You remember the original announcement of expert integrated systems, Mills talked about the IT labor problem, like two thirds of the cost and servers and storage and related infrastructure as people. And you're attacking that problem. What, Peter, my question is what kind of operational changes are your customers making to accommodate Pure? Because, well, it sounds good, sounds great actually. A lot of customers have different cycles and they get different lease, they got a three year lease and it's up this next year on storage, the year after on servers. So how are they rationalizing those discontinuities? Well, it's interesting that you bring up the organization dynamic, because it is interesting for many of our clients. If you think about it today, they've been in a model where they literally have separate technology silos. They build their skills around those silos, they're buying cycles around those silos, and then they have the job of integrating it all together and making it work. And they develop deep specialty skills in there as well. Now, when you introduce a system like this, where a lot of that integration is done, a lot of the technology is built in, it really is above those cuts across those silos, I should say. So it does create an interesting organizational dynamic. And in many cases, we find ourselves articulating this value proposition to the CIO who may be in the midst of trying to understand how to transform their organization. But what we also find is that those skill sets, they see the value in this. They don't want to spend all of their time tuning. They really don't. They much rather get onto that next application, providing that next level of function, exploiting that next application, whether it's a mobile-driven solution or trying to get their services. So they spend far too much time on the tuning and the integration themselves. So what we're actually finding is organizationally, a lot of times, different roles get together and rally around this type of technology. Now, there's typically an application or a use case that brings it in. So for example, in the case of our pure application system, we find that clients will start with their application dev test environment. So you'll have a piece of the organization that provides middleware services today. And so you can bring in an integrated solution like this. They can begin to get comfortable with this idea of patterns. By the way, IBM delivers patterns. Our partners deliver patterns, but clients can build their own patterns as well based on their own application portfolio and their own best practices. So that's exciting for them to get started on. And so there's typically certain application scenarios that are a good starting point for them. How do you recommend, Peter, given that there's some organizational issues, how do you recommend customers go from point A to point B to get to some of those siloed systems to this new, more streamlined environment? Well, in some cases, and I think I'm losing my microphone here. In some cases, again, it's that application starting point. In the case of- Pick a spot. Pure data, pick a spot. So pure data for analytics is an example, is that data warehouse environment, that high-speed, high-performance data warehousing environment. So you pick a spot and you can go in and you can drop in that solution. In other cases, they may be experiencing a pain point with their existing environment. So there's an opportunity to consolidate, let's say, lots and lots of databases onto a pure data for transaction system. So there's a consolidation value proposition. They may be having application performance issues today. It actually kind of shows up two ways. Clients are typically either trying to reduce their operating expense. So that's one side of the coin. There, it's about consolidation or optimizing for performance. The other side of the coin is to accelerate deployment of that next application. Mobile's a big driver. Cloud-based services are a big driver. So that's the other side of the coin. So they're either cutting costs or optimizing or deploying that next application. And we have clients that are starting in your place. So I had, at the pure launch, I had the pleasure of having lunch with Steve Mills and I was having a nice conversation with him. I said, what's the premium that customers have to pay to get pure systems, expert integrated systems, relative to rolling your own? And he said, I do all the pricing. I review all the pricing for every one of these products. I was going to ask you what he said. So I know what I'm talking about. And he said, there is no premium. I said, I don't buy it. Why aren't you charging a premium? He goes, that's what we chose to do. We are trying to grow this market so we are not charging a premium. So then I followed up the obvious questions. Then why wouldn't anybody do it? Why wouldn't everybody do this? He said, because they're afraid of getting locked in. He was very candid about that. So I'm sure you hear that a lot. You know, a lot of people have said this is the mother of all lock-ins because it's very alluring. You think of these blocks of infrastructure. What do you tell customers to say, Peter, I'm afraid you're going to give me this great deal, this concierge service, all this wonderful infrastructure and application knowledge. And then I'm afraid you're going to jack the price down the road. What do you tell those guys? Well, first of all, IBM's had a long history of protecting our clients' investments. So when we invest in something, we're in it for the long haul. It's not like we're going to stop support of a particular application environment when we protect our customers' investments. In this case, what was very important to our clients was not so much the ingredients that we used to build the systems, but that we were still preserving their choice. And that choice just shows up in different ways. Could be the choice of the applications that they run in the environment. So for example, with our PureFlex system, PureFlex is very much an infrastructure system. So it's just a cloud infrastructure. Yeah, it's a cloud infrastructure, but you have a lot of flexibility, hence the name. You can run a Unix-based Power environment, an X86 environment, a combination thereof, all fully virtualized choice of virtualization, hypervisors. So we provide a lot of flexibility in terms of how you take that box, essentially. And you can then deploy any application or middleware you want on it, including our competitors. So the network is fixed, right? That's your own design. The network within the box, right? Yeah, sure, the external to the box, as you choose whatever you want. Exactly. But within the box, which is a good thing. You guys re-architected it from the ground up. Okay, so I can attach virtually anybody's storage on to it, right? We can, in the case of the Pureflex environment, the answer's a little different depending on the machine. So with the Pureflex, you can come with the V7000, which by the way is a fully virtualized storage environment. And it allows you to not only virtualize the V7000 storage, but it'll allow you to virtualize other vendors' storage. Right, so I can hang pretty much anybody's off of it. I've got to have that core system. You do. But that's a prerequisite, but once I have that, I can hang anybody else's off of it. So we preserve choice where it's important to the client. And so, okay, and then, but it's IBM servers or? It's IBM Server Technology. So the network internal and the IBM servers are fixed. But in the storage base, but you can hang other storage off, which is unique. Right. I don't think anybody else does it. Maybe Hitachi does that, but nobody else does that. And maybe with some clue. You have a choice of virtualization. Virtualization technologies and the like. And then we have an integrated management approach. It allows you to manage all of the elements of the system, both physical and virtual. But then it plugs into the larger enterprise management software. So again, we're not imposing if you've made an investment in enterprise management software, this thing is not an island to itself. It slots into your design. So I want to finish this line of thought and then John's got some questions. So, but essentially, that gives you a general purpose infrastructure. It is pre-integrated. That's right. As well in your portfolio, you've got location specific infrastructure. And now data. That you're not going to run virtually, your entire cloud application portfolio off of it, but you can focus on your operational analytics. The applications for example. So we have a very broad spectrum of integrated systems. And really the difference is the degree of flexibility that you choose as a client. And so as you move from say an infrastructure solution like Pureflex to a more of an appliance based solution like Pure application of some of the Pure data systems, we're doing a lot of that optimization for you. We're selecting some of that underlying infrastructure because we know how to put it together to drive maximum performance and efficiency for that application. So where you know your application and you want us to optimize for that application, we'll make some of those decisions. Well to me, the bottom line is it expands your total available market as well, because you're playing in all those worlds versus an exadata, which is a very narrow value proposition and can't support the myriad applications across the portfolio, but go ahead, John. I was just saying we're getting the hook here from the producer, so I couldn't get a word in edgewise as Dave and I talk. Never give up the microphone. Dave, you got a good read on Pure, so good job. Thanks for coming on theCUBE. Okay, we'll be back with more back to back coverage, wall to wall coverage here at IBM information on demands. theCUBE, we'll be right back with our next guest. Thank you.