 TheCube at IBM Impact 2014 is brought to you by headline sponsor IBM. Here are your hosts, John Furrier and Paul Gillan. Hey, welcome back everyone. We're here live in Las Vegas for IBM Impact. This is Silicon Angles theCUBE, our flagship program. We go out to the events, we start to sing with the noise. I'm John Furrier, the founder of Silicon Angle. Here's my co-host, Paul Gillan, co-hosting here with Silicon Angle. And our next guest, we're excited to have Steve Mills, Senior Vice President, Group Executive, IBM Software and Systems. Welcome back, Steve, to theCUBE again. Good to be here. Great to have you. I got to say, the Pulse interview was fantastic. We had so many tweetable moments, quite a great interview. It was setting the table, really, for what's happening here with customers. So, first take on Impact here has been about power systems, blue mix kind of getting some legs. Sea legs are kind of getting established. Is that what you guys expect? Is that what you guys wanted to say here? And what's the key messaging that you guys are putting forth? Well, I think clearly those are two of the most important focus areas for the conference. In our blue mix, we launched at Pulse and it went into beta. We got the, actually many thousands of developers to begin to use it. We're now kind of ready to come out into the marketplace in a more formal way. And blue mixes, as we've discussed, it's an open platform for DevOps. It's all about APIs and components, how to rapidly develop applications. And it's sort of these next generation applications that are very much around engagement, interaction. It's certainly not the kind of thing you look to for your back end ledger based systems, but it's all about those end user facing and customer facing environments where you need that kind of rapid development, iteration, mobile device enablement, those kinds of things. And it's where a lot of the excitement is in the industry today. So blue mix, open platform, but with back end integration into your back end environments and very much geared towards not just what the corporate developer might be looking for, but also the new business startup. The kind of thing where entrepreneurs can come in and find componentry that can map against the kinds of things that they want to do. It's interesting that DevOps certainly captured the attention of the early adopters around Amazon and the public cloud. But when you start talking about enterprise, it's not just the start of getting a little more resource. It's enterprise grades, serious conversations. So you've got to please two masters, the in-house kind of nuts and bolts in the kitchen, so to speak, and building out and growing the infrastructure capabilities. And then the business line manager you mentioned, which are basically like, hey, just get the job done. So is that something that you see? Is it two masters? Is it designed that way? How do you view that? Is that a misread on my part? No, I think that for whether it's a corporate in-house application development team or somebody who's trying to build up a new capability for commercial sale, you do end up serving multiple masters. If your end target audience is going to be a larger-sized business, you're going to deal with, on the one hand, can I get the rapid development and deployment capability that I've come to expect around these DevOps environments. And by the way, I've got to fit into the corporate infrastructure. Do I comply with that? So if I'm an entrepreneur starting a company and my customer base is going to be enterprises, this environment provides me with ways to create those links into traditional enterprise backends. So I arrive not just with the next SaaS offering that stands alone in the cloud, but it's a SaaS offering that integrates better into the corporate customer environment. And let's face it, lots of entrepreneurs are starting companies where their target buyer is a corporate buyer. You've got, of course, you're going up directly against Amazon now, you've been targeting Amazon, you're advertising, it's very clear that they're the bad guys right now. Amazon's buzz has been sort of around, they're the default for the startup company, for the new company to go put everything in the cloud on Amazon. IBM has legacy of the enterprise customers. How do you make those enterprise customers that legacy into an asset instead of liability? Well, a couple of things. First of all, just to sort of correct the record, the SoftLayers company, we bought them last year, there are more URLs running on SoftLayers and on Amazon AWS. So in fact, there's a very popular environment for startups. I think people misjudge that or have bad information. SoftLayers is more popular dramatically so than Google and it's hosting more URLs than any other software infrastructure environment in the industry today. So we start out with a good base with all the classical tools and techniques. Get on with a credit card, rapid deployment, now we're adding platform as a service, dev ops. We're bringing our SaaS properties on, we have over 100 SaaS offerings in our portfolio that'll be up and running on SoftLayers. And then of course you get to the enterprise buyer who in many cases has been sandboxing in AWS or SoftLayers or somewhere else but is often struggled with how does that translate then into production. So I might do some development work out in the cloud but for reasons of corporate rules and regulations I'm not putting production. And the SoftLayers environment provides a very unique infrastructure, 1600 APIs, rendered explicit, you can see absolutely everything going on, it maps to the corporate audit and compliance requirements. We're incorporating security services, data loss protection, integration, compliance with different standards. So all of this is geared towards satisfying the needs and requirements of the corporate world as well as the startup world. Well that's interesting, audit's a great example. I mean for instance Amazon's not strong, I don't think they have audits or at this point but that's a great example, that's table stakes, right? You can't roll into the enterprise without audits. Do you see that? Those kinds of things, the real kind of straw that breaks the camel's back to. Well I think it's absolutely a requirement for corporations to go into production. And again it's providing 100% visibility. Can I see what server I'm running on at the serial number level? Can I see all of the storage volumes that are being used for the data? Can I see all the changes, the updates so I have a total 100% auditable path for how that environment's been operating and everything that has happened to it. If it's 100% traceable, I now have the ability to satisfy my auditors that I've got a reliable environment that they can actually interrogate and understand what's happening. Is that just, Amazon's just not ready there, they're growing up and they're still kind of like in elementary school for enterprises or is it just that just haven't gotten to it yet or is it a technical limitation from your thinking? I think putting aside for a moment what they have gotten to, in fact the matter is they haven't necessarily staked out that production environment as being their top environment, I mean their last few years obviously they've had a lot of success in attracting start-ups. They've done well in the public cloud. The venture funded companies find Amazon AWS to be a place to start because people's colleagues they're starting there so they're attracted to that. The corporate environment has a lot of other requirements and frankly Amazon will have to do a lot of evolution of their platform to get up to the level of auditability that we have. We know they're working on it. Wouldn't you like to reach some of those and wouldn't you like to have a Netflix in your portfolio? How do you, what's the strategy of getting to those companies? Well, obviously you have to provide better economics, better manageability and quality of service that makes your environment attractive. So we have a lower, as the amount of data that you need to store away increases our costs remain relatively linear and AWS costs skyrocket. So for large amounts of data and in fact matter is in all of these public cloud environments if the only thing you're looking at is cost to compute you're not looking at the real cost because the real cost is networking and IO. And as the data increases the soft layers environment actually becomes more economical than the Amazon environment. Needless to say we're out talking to a lot of companies about their options of moving on to soft layers. By the way, here at this event we're featuring the Aspera technology. Aspera is the underlying high performance, high bandwidth, point to point and multicast technology that underpins Netflix, right? It's the mechanism they use to deliver their streams. Well, we acquired the company. So they're here at the event and we're showing off this capability. Hey, the tail can wag the dog as they say. You know, so to speak. So look, we're hunting out where the profit pools are. So it's more than just having a lot of customers or, you know, having a lot of volume. It's also, where can you make the money? I love talking to you. I like how you talk. Sticking claim is a good way to look at things. So before we talk about power for a second I want to talk about the cloud marketplace, right? So as you stake out that claim it's not just an app store. It's not just like the flavor of the month. Talk about what you mean by cloud marketplace. Is it for businesses? Is it for the developers both? What is the clean positioning of cloud marketplace? Well, where we're going with marketplace is to make available a very broad range of components and API interfaces to invoke function. And this is something that we'll populate with components. We're going to work with various companies, systems integrators and ISVs to encourage them to push their components onto the environment. We're going to let our corporate customers actually create their own private development instance within the blue mix infrastructure. And then they'll use that marketplace to get at components that they use. And by the way, they'll be able to put their own components in but keep them closed. In other words, not open them up. You have a private marketplace. Yeah, so they have a little private marketplace for themselves. And everybody knows that the greatest efficiency comes through high level of reuse, right less code, reuse the components you have, taking advantage of as many API based interface structures and things that you can pull in. And these very highly productive DevOps models are very much based upon this whole idea of code less, design more, script, use components, bring things in that are already built. And also distribution, developers get distribution and sharing, so that brings that openness feel back to it. Is that part of the design as well? Absolutely part of the design. So it's, what people will find there is not just things from IBM but from every other part of the industry. Obviously, open source components, things that are available broadly in the marketplace on some of the other environments that are out there. You're going to find those in the IBM environment. So, you think about what people have found and liked about AWS. Think about what they've found and liked about Azure, Heroku, Engine Yard. Go through the list of all those players. All of that and more will be in the IBM marketplace. Now, Steve, marketplace is as Apple and Google and Salesforce and such have defined very much a numbers game. It's how many partners you get in there, how many customers you get in there. How should we evaluate the success of marketplace a year hence? Well, we're going to provide that score sheet of value. So what starts out as hundreds becomes thousands and then tens of thousands. Obviously, at some point now the richness of the environment is less the issue because you can find a great many things there and then it gets into aspects of tooling, patterns. Particularly for many developers, they're looking for a pattern to be instantiated in the environment. Let me show you what others have done. Here's a sample set, here's a pattern. If you want to build this type of application, here are the things that you bring together in order to effectively turn it into a kit that delivers the application that you want. But clearly, volume, volume, volume counts in the beginning. Get as many components as you possibly can into the environment and then begin to create these sample sets, the patterns, the pre-kitted, if you will. All right, so it's great to be able to find lots of things. It's even better to be able to find exactly what you want. Steve, I got to ask you about the bubble. I wrote a couple of articles this week about it. Obviously, we saw a couple of things. Intel made a big acquisition, I mean, investment in Cloudera. Huge valuation over the top. Pure storage, which is a storage array. 25% the value of NetApp, I mean, $3 billion. I mean, it's frothy, but it's an innovation bubble. What's your take on this environment right now? I mean, obviously, valuations are high. Is this an indication of the old storage bubble? Is it the convergence? Is it a new animal we've never seen before? I mean, what's your take on that? Well, look, I think technology generates a lot of excitement. You know, I mean, the information technology is the transforming technology, certainly of our age and will continue to be, I think, you know, throughout the rest of this century. And so that creates excitement and it creates a sense of future value. Now, at some point, you look at those valuations and you say, well, how does anybody get a return on that? But clearly, if someone's willing to spend the money, you sort of say, well, you know, it's kind of like your house. It might not be as nice as your neighbor's house, but if someone comes in and really wants it and wants to give you a premium, why wouldn't you take it type of thing? I mean, Fusion IOs is looking pretty good right now. So is Violin Memory. It's $300 million valuation. Did you say they overpaid? Well, the frothiness, it's not sustainable, right? You know, it's not sustainable forever, right? So there, you get these peaks and valleys. You know, WhatsApp, does that price make any sense? Only in light of how much money, I guess Facebook had had an equity to spend with it. And by the way, you should know that WhatsApp runs on softwares. Yeah, we nailed that at pulse. Okay, I mean, you've done a lot of small acquisitions, very large number of small acquisitions and by all accounts, a good job of integrating them. How about a big acquisition? What kind of latitude do you have to do a blockbuster? Well, it doesn't really match the model that we're on. I want to be able to take advantage of the skills, the talent, the energy of the companies I acquire without having to spend a lot of time dealing with aspects of culture. And the larger the company, the more the cultural issues begin to weigh you down and slow down the process. We have a huge portfolio. This is, in fact, the largest software portfolio in the tech industry. As big as it is, there are always gaps. If I can find those gap fillers, then what I can do with those gap fillers is actually put them in my go-to-market capability. I'm in 170 countries with tens of thousands of people representing IBM. I can scale up those companies dramatically faster than they could ever scale on their own. And the IBM shareholder gets a nice return on that investment, right? Because that investment of IBM Cash becomes accreted very quickly. The larger the acquisition, the harder it is to run that play. We've done some, you know, clearly some larger ones, billions of dollars, but you know, relative to our size, the actual size of the company. 10 billion is a new one billion, they've been saying. No, at some point, it doesn't make sense to shareholders. You guys have been good at tuck-unders. You guys have, you guys pick good value deals, make them part of the engine. That's what you're saying, is you're saying, pick the risk up the table. It's our shareholder's money. You know, we work for the people that own us. It's their money. If you're an investor in IBM, you want to see IBM use the money responsibly. And the last thing you ever want to see is an autonomy-like acquisition, you know, coming from IBM. That's not on our menu. No way, no how. On the white spaces now, obviously being, you guys have a lot of organic R&D going on. So, and also you got the M&A, which is kind of the tuck-under strategies we just talked about. What are the white spaces that you see out there that you see it obvious, you know, to be, you know, determined, you know, build out areas that you're going to fill in with organic and possibly some M&A deals. Is it the orchestration? Is it automation? What are you looking at as territories that you need to claim in there? Well, first of all, we look across our whole portfolio as to the kinds of things that we want to go after. And, you know, it's understanding the patterns of implementation of the products we have. Where do those gaps exist? And would a piece of technology from another company fill in that gap? Very often we're partnering with the companies that we buy before we buy them. So we see how they fill that gap in what the leverage is. And we're investing obviously in the hot areas. So, you know, you think about, you know, mobile security, all kinds of analytics related acquisitions, things that fit into the cloud, which is more about deployment. So, you know, we're investing where the market is moving the fastest. Because that's where the greatest pressure is to, you know, try to keep up with where the market's going. We did set out to talk about power. So we should talk a little bit about power before we're done here. Open power, you've got a huge endorsement from Google today, so it's moving its stack to open power. However, it occurs to me, you're taking on Intel, essentially you're competing with Intel now, at a time when you have a lot of other initiatives to pay attention to. Why is the open power initiative not a distraction from other bigger prizes you could be going after? Well, power is a big business for us. There are 100,000 plus companies out there that use power. It's an important franchise. And we're looking at it from a long-term perspective. The market, I think, has been looking for server alternatives. You know, clearly the server market space has compressed as far as players are concerned. IBM and Intel are viewed as kind of the last two companies standing. And clearly Intel has volume advantages. No question about it. You know, so we're looking at our position in the marketplace and what companies like Google and others are looking for to balance out their server portfolio. What attracted them to power was the fact that this is a very high-performing chip. It's got excellent threading characteristics, data throughput. So the amount of compute power per unit of electricity consumed and heat output is outstanding. You then combine that with an open licensing model, which is what we adopted around open power. We essentially looked at the kind of licensing patterns that ARM, for example, has benefited from, which is the popular chip in handheld devices. And so we've gone out with this ARM-style licensing. That means that for relatively low entry point cost, you become part of open power. You can source from us. You can actually fabricate with others so you're not uniquely tied to IBM. You can find others that can make these kinds of things. You can do derivative works off of your license. So it's a non-restrictive license. It's not a proprietary license. It's open license. It's derivative works, but how is this different from a GNU public license, that type of model, where your derivative works become part of the open stack? Well, we're not using a GPL-style licensing, right? So if you want to do extensions, you can do extensions, and those are your extensions. So you think of Apache-style licensing in the software world, you can actually leverage our license in the same way you would. You don't have to contribute the IP back to the... But you own the extensions you create. You own the extensions you create. Oh, by the way, as a participant in the community, you can also offer those extensions back. So there you have choice. It's not dissimilar from other, if you will, non-GPL-like licensing models, where you have the freedom of action. But as a community participant, you may very well want to push it back in. It's personal innovation. Is there another shoe to drop in the Google relationship? They have enabled their stack on power, and they obviously, they acquire chips. They go to third-party fabricators, and they make the sleds that shove into their racks, and power will now be part of their configuration. Steve, what's your big bet on the power? I mean, obviously you're seeing some trends that we were speculating earlier on. Obviously we're going into a maker movement-like developer market where you've got tinkerers playing with hardware and software, so the threading thing makes a lot of sense there, but also the analytics market with internet of things is exploding. Obviously you're ingesting unstructured data and just dealing with it later is a big part of the analytics piece. Did you see those things? Is that one of the big, what's your big bet of those? Well, one of the key design points of power eight and the power of family going forward is what we refer to as this high-performance bus architecture that allows essentially a coherent programming environment, but you can connect other processors into power. So I program through the power chip, that's what my developers see, but I can actually schedule a workout to a field programmable Gatorade, to a GPU, so we're partnering with NVIDIA, Altera, Xilinx, the popular providers, and therefore we're allowing those that want to engineer unique system structures to be able to do that. So if I wanted unique encryption to improve my security, I could use an FPGA and do custom encryption. If I wanted to do some kind of advanced analytics, Hadoop related speed-ups, I want to do pre-processing around predicates and sorts and those things. I could be using the NVIDIA GPU program natively through the power chip, but get all these other advantages. And so we've made this an open environment that you can plug other things in. So in a sense, you can build your own hot rod. I mean, this is all part of the big development, right? I know, I love it. Everybody wants to build their own hot rod. And if the pieces and the components are there and you have good engineering skills, you can build your own hot rod around this. It's like American graffiti for technology, just cruising down Main Street with your hot rod. Absolutely. So what does your hot rod need to do? Is it big data, big transactions, low latency, data encryption, you talk about the domain and we can adapt to that domain. So you get a souped up device or car with the engine you want under the hood. But what's interesting too, would it be safe to say in the statement that with power systems, the big bet is that you can provide high performance technology at the same time support scale out open source commodity hardware. Exactly. So that's essentially the bet, right? Yeah. You're not going to pick a side. You can win in both. We moved the architecture to a little Indian. We've made Linux 100% portable without change across into the power environment. We make it easy for all that work to come across. We have this hot rodding capability around the architecture, open licensing. So it's multiple dimensions in which the leverage, the base level capability of power, which has always gotten high marks for its scalability characteristics. You bring to the common man now. And now make it possible for other engineers to jump in and do their own thing around this environment. And frankly, that's what many people in the industry want to do. They want to do their own thing. Steve, anything in the Watson area specific to power? Will we see kind of a Watson hot rod that IBM might build? Well, Watson runs on Linux on power today, where all of the Watson deployments we're doing are being done on power. Right, but optimized power is what I mean. And with power eight, we already have a level of optimization through power eight's unique features, for high bandwidth IO and threading. And then we're actually doing a set of customizations, specifically for Watson to take advantage of the NVIDIA GPUs and how they can speed up some of that work that the cognitive system needs to do. So they'll be a Watson hot rod. Okay, but Judith Hurwitz was on earlier talking about Watson and saying that Watson is such a complicated ecosystem. There's so much to Watson that it's, IBM has to move slowly and moving it out there really as a general purpose engine. Are we going to see, we've seen a lot in healthcare lately. Are we going to see other industries go big in Watson in the near future? Yeah, and we've already had some announcements in financial services, insurance, advisory type services, the things that would be clearly a bit simpler than the healthcare challenge that we're taking on with Watson. You know, we have announced at this event a Watson environment for corporate developers to actually come in and sandbox. We're encouraging entrepreneurs. We've got a dozen different universities that beginning their next fall cycle begin to teach the principles and structures around the Watson environment. So we are clearly opening up to a broader ecosystem and we've made it modular. It's a system, but we modularized around a platform structure so you can effectively only use the pieces that matter to your particular application scenario. So Steve, if I want to shift gears into big data, but before we do that, I want to let you know that you're famous on Twitter because you now have a fake Steve Mills Twitter. Did you see that? I've seen this before. I don't know who this individual is, but they... They're certainly not doing a good job being the fake Steve Mills. They do seem to want to take advantage of... I'll see you in Twitter. It's a compliment. We consider it a compliment. But that brings up social data as a form of big data. So I got to ask you, do you see data as just a part of the fabric of the platform across the portfolio? Or do you see it becoming a P&L? Because we've talked to folks today from power to streaming web sphere across the board. It's all big data nuggets sprinkled into the platform. Is it going to be just an embedded feature across the portfolio or is there a P&L around it? First of all, we do measure our total business analytics business. Bob Pitchiano, who works for me, has responsibility for all the basic material that makes up our business and analytics portfolio. And that includes those things that apply to big data scenarios. So that's solutions. Yes, now frankly, I haven't met anybody recently that's doing little data. So every analytics project that I'm touching is always big. And it's how many terabytes off it in the hundreds on the way to petabyte scale environment. So I do think that the market has moved in that direction because the price points have come down, the ability to ingest and analyze more data has improved economically. And so businesses are saying, well, if I can bring in more data, can I find more patterns? Can I unearth more nuggets, more jewels? Things I could act upon that previously I couldn't act upon because I could never afford to bring all that much data together. So frankly, all these analytics projects have become big data. It's hard to nail the technology you're saying into one division, but the solutions you certainly can P&L price out and sell to. Absolutely, yeah. And frankly, it's mixed data types. It's structured and unstructured. And classical relational fits, but certainly not the only technology. And you've got to deal with document format data, graph data, object-oriented structure. So many different data types and techniques. We have a very broad portfolio that touches on all these different mechanisms. We're a Hadoop distributor. You saw we acquired Cloudant recently. We're doing a lot of things around very complex data structures in many industries. Is big data a P&L for you, or is it an app? Well, business analytics and database are P&L structures for us. We don't explicitly break out big data as a unique space just because- It's broad. What happens is everything is quickly big data. Again, who's doing small data? All data, all data is big data, yeah. That's just the way of the world. That's where the world's going. We were joking earlier about back to the 80s as we were talking about some of the, how the mainframe selling was 50th anniversary, the innovation around that. And so just from the terminology we've been kicking around today, data processing, decision support systems, MIFS, it feels a lot like those days of, you know, when that kind of language is being used. Now you're hearing it mainstream, pipelining data in this new era of Cloud. So is Cloud the mainframe, has always been the discussion, how does that look like? I'm not saying the mainframe for IBM, but like as a metaphor, it seems like a mainframe, I mean power systems, you're essentially saying, here's some high performance capacity in brains to whatever you're connected to, here's a marketplace, it feels like a mainframe. Do you guys? Yeah, we'll see how these things evolve. I think that we're really in many ways at the early stage around these public cloud environments. Now, there are numerous companies measured in the tens of thousands that have been delivering services electronically, and they're often production services, and by the way, mainframes are part of their environment. So if you go into ADP, Depository Trust Company, go into United Parcel Service, anybody that's put up electronic services around their capability, you'll often find mainframes involved in that service offering. Their customers are interested in the outcome, they're less interested in what the underlying technology is, but as a provider, they've built up an incredible reputation for reliability, availability, security, and so on, they're trusted providers. The public cloud environment has obviously been characterized by startup companies, entrepreneurial activity, a lot of systems of engagement, types of approaches, new apps that are touching users, not nearly as much on the production side, but this public cloud world will grow up to deliver more production ready infrastructure, and some of those mainframe attributes are going to have to be evident in the environment for reliability, availability, failover, recovery, et cetera. My final question for you is going to be more about what is the outcome of the cloud, the ultimate determinant of cloud success for all the folks out there. You see Microsoft's doing, they got a war chest of in the billions, some say 30 to 50 billion for M&A activity, the new CEO over there, you got HP, you guys, and everyone else, Amazon. What are going to be the key determining factors for the cloud? Is it going to be economics? Is it going to be scale? What's your take on how you view that chess game? Well, the top 10 items are economics. I mean, you can't deny that it's economics because information technology is a tool that you use to power a business, and so the business model comes first, the technology comes second, and when the economics are right, you can do more with technology. You know, and it's not that other attributes are not important, but the economics take precedent, and the excitement around public cloud is very much driven by the perception that I can find better economics. Can I be more agile? Can I adjust and adapt? Can I get the technology that I need when I need it, but not necessarily have to pay for all of it up front? So all those are appeal points, and they're very much tied to economics, so that is the fundamental driving force. I think one of the mistakes though that is often made in describing the market as it exists today is to ignore the thousands of companies that are business process service companies who predate the use of the term cloud, but who nonetheless are delivering very valuable services electronically, and businesses have for decades been been de-verticalizing in effect. As a company, I might have had all my own systems doing all my own apps 20 years ago, 30 years ago. Today, I use all kinds of providers, and some of them could be literally born on the web. It might very well be salesforce.com for my Salesforce, but it could be ADP for my finance department, and ADP is as much in the cloud as Salesforce is. There is much of dot com as Salesforce is a dot com. You know, it's just a different constituency that's being served by an electronic service. This final question I'd have is how you're helping internally make that switch, because we see a lot in the reseller channel, in the integrator channel, they're trying to move from this box sale compensation structure to selling subscriptions, having a lot of difficulties with their sales force making that transition. What have you been able to do? What kind of success have you had in getting your sales force to transition to selling subscription services? Well, first of all, we did the very simple thing of compensating them equivalent to whether it's on-premise or in SaaS, which means we apply a multiplier that delivers to them compensation. They become neutral to the deployment. That's a good solution. So that's a starting point. Yeah, sales reps are coin operated. If you don't come up with a coin, they're not going to respond. But I think even more important to the point that you're making is that around all of these services, if information technology is something that transforms business, then it's not just the product itself, but rather the roadmap for transformation, which means that there's a services model that goes behind all of these software as a service offerings. How do I get onboarded? How do I adjust my business process to match what the SaaS offering does? So there are big services opportunities, implementation opportunities. Frankly, many of our partners are truthfully regional systems integrators that also carry technology. We're allowing them to participate in the SaaS sale, but in particular, they're engaged in the onboarding and in the enabling of process to actually make the SaaS offering do something for the customer. Steve, thanks so much for joining us. Know you're a press for time. I'll give you the final word. Share with the folks out there the bumper sticker for your customers and your troops internally for your mission over the next year. What's the bumper sticker to the troops and the customers? Well, let me express the bumper sticker in terms of this particular event that we're at because we do a lot of things in IBM. But this event is showing people how this whole idea of rapid, agile, flexible invention innovation can take place within businesses by taking advantage of the kinds of tools and offerings that we're bringing to the marketplace. It's lifting the classical middleware infrastructure environment, if you will, up to this next generation of interfaces, uses, mobility, flexibility, DevOps. It's overlaying the new things into that infrastructure. And what it effectively does is it renews the infrastructure. Now, when you're a 100-year-old company like IBM, you spend a lot of time focusing on how you're being renewal to all the things that you do for businesses. The vitality of our portfolio and the importance of the role we play with our clients is directly related to this idea of renewing, renewing, renewing. And if you can bring that renewal around the latest capability to bear, the customers want to go with you because they've made a big investment over a long period of time. Steve Mills, Senior Vice President here at IBM Edge. I mean, IBM Impact. Be right back with our next guest after this short break.