 Okay, we're back live here in Silicon Angle's theCUBE, our flagship program. We go out to the events and extract a signal from the noise. This is IBM's information on demand live in Las Vegas. This is day one of two days of wall-to-wall coverage. I'm John Furrier, the founder of SiliconAngle.com and I'm joined. I'm Dave Vellante of Wikibon.org and we're here in the exhibitor section, the big data pavilion. This is theCUBE and now's a good time to tell you, go to Wikibon.org, check out all the research, SiliconAngle.com and SiliconAngle TV to check out these videos and check out the content. And we're here with Bernie Spang, who's the director of strategy and marketing for database software and systems for the information management group at IBM. Bernie, welcome to theCUBE. Thank you, it's great to be here. So John, we've been talking today and recently about how database was sort of a boring industry for a while and now it's the hottest thing going. Yeah, I got one of my tracks in my computer science degree in the late 80s was in database and operating systems. I passed on the compiler track, but just even three years ago, you went at a cocktail party, hey, what do you do? Oh yeah, I'm a database guy. Yes, kind of people don't really fall out of their chair when they hear that, but today it's the hottest thing on the planet. Databases are changing. Everything about big data is the key element. It's a key ingredient for what's happening, whether you're talking about Hadoop and HBase, anything on the open source side, all the way up to DB2 at IBM and everything in between. So Bernie, that's in your wheelhouse. It is. So tell us, it's cool to be a database guy now. I'll tell you, hearing you tell that sort of reminded me when I moved to the information management team from WebSphere and I was involved in the early days of Java and Eclipse launch was one of my things. People say, why are you going to database software? That is boring. Yeah, it's commodity stuff. And ever since I've been here, it's been anything but. And that's largely driven by the convergence of a number of factors. The amount of information available that the internet has brought to bear, especially now with mobile devices, really expanding that. The sheer compute power that is now cost effective to be able to apply to that information and then the advances in analytics technology. Solid state drives have a big part of that too. Well, that's a big part of that computational foundation making it cost effective. Now it's more cost effective to keep that information in flash memory. Just DRAM memory itself. You know, you could have more cost effective rate and then just the processor power. But then you apply, you add that to, or you add to that the software advances. The new techniques for analyzing information and being able to analyze not just structured data in a relational warehouse, but also the unstructured information and the text and the video, et cetera. And with our streams capability, being able to analyze information in motion, that's a whole new area. Take us through some of the dynamics in the marketplace. We'll start in kindergarten, work our way up to college in the conversation. Explaining what's going on in the dynamics of the database market right now, relative to the tech speak, meaning what's happening, what's the key drivers around, old way, new way, new guys coming in the market, new CS degrees, people who are on play with database, they get the old school who are locked in their way. So there's a little bit of a conflict there. We see that with DevOps and developers and operations in the cloud. But in the database world, you've got kind of the old BI, data warehouse guys, hardcore database, and then you got the new school who just wants things to move fast. Well, I mean, this happens in every generation of technology. I mean, I've been with IBM enough years that I saw it with the mainframe and their system Z evolution, oh, the personal computers and minis come out, therefore everything should go that way. And the days of the dinosaur discussions, meanwhile, we're growing the horsepower on the mainframes, leaps and bounds every year over year. The point I'm making is new technology comes out that does new things or does some old things even better, that does not mean it solves all problems. And so that's really the story here. When you say, you know, what's going on with the database, the traditional relational database market, there's really two things. First, there's a recognition that relational database technology is not the right best fit for every data challenge. And that's where you get things like the new kinds of technologies, graph stores, Hadoop, time series data management like we do with our Informix software is being a completely different approach. So that's one thing, the recognition that you want to apply the right best tool for each job. Then in the case where the right best tool is a relational data system because that still is the best answer for a lot of data management workloads, then it becomes cause, how do I do that more cost effectively, simply and with greater performance? So those basics have not gone away. So I'm inferring from that Bernie that IBM strategy is Horses for Courses. You've got a rich portfolio, a DB2, IMS, you mentioned Informix, others. Talk about the portfolio generally, but specifically IBM strategy. So you're right, the Horses for Courses is a good expression for what we mean. I like to say, there are others in the market who suggest that they offer a one-size-fits-all solution. And I'm quick to remind folks, no, what they're offering is a one-size-does-not-best-fit-all solution. They're asking their customers to make a compromise, either in efficiency or performance or cost or complexity of using that one thing to try to do multiple different things. So our strategy is have the right best tool for each job and make sure that they all work together. And especially in the information world, a big focus of our InfoSphere portfolio is managing information integration and governance. Now that can't just be for structured data in a relational system, it's got to be across all systems, right? And the flow of information and the governance and the security of information across all your different types of systems. So many have sort of positioned Hadoop as the batch and many of the traditional companies that said come to us for real time. What's IBM's angle on the whole unification of the sequel and the non-sequel worlds and how is that coming together in your customer base? So definitely we're seeing an increased interest in real time or operational use of information, operational analytics. One of the big exciting pieces of news here this week is our pure data system announcement. And one of the models, now getting back to our strategy, it's available in three different models. A pure data system for transactions, for analytics and for operational analytics. And that operational analytics capability is there so that you can do the deep analytics in a warehouse but also be able to access that information and tap into it for real time decision making. Another example is the InfoSphere Streams capability that's actually analyzing information as it flows through the system before it's even stored. Or even if it's never stored because it's just so much information. So there's again a spectrum of opportunities to analyze information in motion over time when it's stored, do long running analytics but also immediate operational analytics. I have to give you a props. I mean IBM does a very good job of sort of hiding the complexities of the technology behind its good names, good branding, but so but the transactional stuff would be the traditional DB2, would it not? The pure data system for transactions is DB2. And then a TISA piece fits in, talk about that. With the pure scale capability of DB2. So it's a scalable database cluster that's transparent to the application. The pure data system for analytics is powered by a TISA technology. And this is the system that is the evolution of the NETISA platform that used to be called TwinFin before the acquisition. And that's really focused on specifically on the deep analytics workload where you want to optimize and simplify that. And then the pure data system for operational analytics is built with the InfoSphere warehouse technology which is also DB2 at its core. And that enables the support of continuous ingest of data, continuous loading of data. The deep analytics and supports more than a thousand operational accesses to the warehouse to support those operational analytic workloads where you want to do real time decision making. So architecturally, you've preserved the essence of those architectures, if I understand it, but the package looks a lot different. You're bringing together storage and servers and networking into more of an appliance-like model. Exactly. So the pure data systems are the latest of the pure systems family that we launched in April. So what we refer to as expert integrated systems. So the characteristics of what we mean by that is built-in expertise, integration by design that's optimized for the target workload, and a simplified experience throughout the life cycle of that system. So each of these systems integrates the compute, storage, memory, system management workload, provides a single supportive maintenance and support for the entire system, right? And a consistency across the portfolio to do what you just said. So your organizationally, your customers are now faced with some dislocations, right? They've got, you're talking about, at least from a branding standpoint, analytics and transactional systems, and you're also talking about from an infrastructure standpoint, bringing together a lot of different disciplines. How are your clients dealing with those organizational dislocations? Well, it definitely comes up. There's definitely a mind shift that has to take place when you're talking about buying integrated systems that have that appliance simplicity in an organization that's used to buying components. You know, I buy the software, I buy the servers, I buy the storage, I buy the networking, all separate teams and put it together. But it's what I just went through, that litany of I got to design the system with all those components, I have to procure it, install it, configure it, test it, and all the experts that have to do that. That represents a huge amount of time and cost and risk of human error that our clients face every single system they put together. And so by using this expert integrated approach, you can save time, save money, reduce the risk, and shift that expense to an investment in new growth projects like we're talking about here with how do I find the resources to tap into this big data opportunity when I'm spending all my money on operations while you greatly simplify the operations, make it more efficient so that you can make that investment. So the presidential debate tonight, I don't want your opinion on the politics, but of course one of the things Republicans, Democrats do, they say, all right, look it, we stand for this versus we stand for that, they try to create these differences. So one of your large competitors says, big data meet big iron, you alluded to the fact that I'll say, you were talking about Oracle, says one size fits all, and by the way, that's for transactional simplicity. I don't mean the transaction of the execution of the code, I mean the transaction of the signing of the check. But so how are you different than the competition? I wonder if you could talk about that a little bit and just paint a picture there. Certainly. So the biggest thing that stands out for me is the workload optimized strategy, right? Versus the one size does not best fit all strategy. And we are focused on helping our clients deliver business value, right? I mean the name of the company is, you know, International Business Machines, that's what we are at our core, is helping our clients with their business results. And so in today's era of computing and the big data challenges and opportunities have only accentuated this opportunity, there's a great opportunity to simplify the IT infrastructure. And so if our clients allow us to do the work, to design and configure and tune and have systems that come factory ready, in fact, you know, to do specific workloads, that frees up their time and energy and they don't have to customize those things themselves. So I wonder if I could follow up on that, because what you said sounds great. We're all about business value. Who doesn't want business value? But customers will often say, okay, but I'm afraid that that means that's a euphemism for IBM services. Now you're embedding a lot of these services and knowledge, service knowledge, application knowledge into these systems. That's right. Talk about what the premium is that a customer has to pay for that and is it still a services led engagement? So no, when we talk about the pure systems, these are systems that have the expertise built in. Now the expertise comes from decades of working with thousands of clients around the world, from our services teams, but also from our lab teams, our development teams, and IBM research, for that matter. And your customers, I would imagine. Right, working with the customers. So these systems aren't just bundles or collections of things that we put together. They actually, the configurations, the automated deployment and management is based on patterns that we've seen over and over again. And we've captured that expertise and built it in so that the clients don't either have to do it themselves or hire IBM or any other service provider. Bernie, one of the things we've been talking about in theCUBE, and it's looking angle, Wikibon in the community is, we've seen this movie before. Specialism around the database moves to general purpose, back to relational, swings back to another cycle. theCUBE, a whole lab cube. Objects in relation, back to relational, more general purpose. Objects, we've seen it, now we've got no sequel. So we're in that specialty phase. And so it's not a mutually exclusive situation, you mentioned that. So where are we in that movement of the market shift? Obviously, you see in Duke World this week, we'll be there covering it. A lot of startups are coming out. You see, Taylor made suits, if you will, of solutions using these purpose-built or these specialty products. But to go mass adoption, you need to have kind of a general purpose. Well, yeah, and actually what you need is, well, when you have new technologies and no sequel, not only sequel, so let's not say it's not sequel, but things like the graph stores, the Hadoop approach, and I'll come back to the time series example in a minute is a great one. When application developers and solution providers see a new technology or a different technology that gives them a compelling advantage, compelling performance, simplicity, or efficiency, you can't ignore that, you got to go after that. Now, in order to gain that value, you then have to create applications or tailor applications to that new technology. And that's an investment. That's not an automatic freebie. So none of these new technologies are a magic silver bullet that cover everything and none of them come immediately for free. But if the value is significant enough, the investment's made. So let me come back to the Informix time series one. Informix has had time series data management capability for years now, but it really has been untapped until the smart meters and sensors generating all this data really started to come into the focus with all these smarter planets, smarter energy, smarter cities, smarter transportation. So when you have time stamped data coming from these devices and you put it in a relational data table, it's highly inefficient. It takes too long to load, too long to process, uses too much storage space. If you use a time series data structure, can use as little as one third the storage, you know, your data load times go from hours to minutes, if not second. Now the beauty of Informix software is it has that data model in it as well as the relational, so you get a benefit of both. But it's a no sequel technology. It's not standard relational sequel. And so we've got new partners now and new customers, especially in energy and utilities. I was talking to somebody from a water utility dealing with all their meters and sensors and not being able to handle that big data challenge. And they're stumped. And we show them what we have here and they go fight and go from data loads that are seven hours down to less than 20 minutes and use one third the amount of storage space. Now that becomes a compelling business value that makes me want to make this shift. I think you're right on the money on that. We're seeing the same thing with Hadoop and HBase for example. Great rows, putting things right next to each other. Low cost commodity storage. So that's phenomenal. The next question I have for you is, okay, what's next after that? So we have that, these databases. Okay, cool, check. Now, a lot of intertwined data. So developers are now looking at data and we coined on theCUBE here that data is a new development kit. So you're seeing developers start to play with data, party with data in a way that's developer centric. See how they've got a co-mingling of data, right? So it's a mashing up data, whatever you want to call it. There's stuff going on now. So what's your view on all that? Where's that tracking to? So a couple of things. First I think having the developers play with data in different ways, like we're talking about here, I think it's going to enable things to happen that we can envision. So that'll be fun to watch. But as these things start to mature and need to be rolling, roll out and become the next mission critical set of capabilities, you know, all the old things come back into play. I got to worry about security. I got to worry about privacy. I got to worry about reliability. It's no more that it's in the sandbox. I need to worry about all those other things. And we're seeing that. I heard one client example, a customer example with our big insights team told me about, you know, there was a Hadoop sandbox project and it had matured and they wanted to bring it in and make it a thing. And then all of a sudden there was a realization of how much sensitive customer information unmasked was on the system. Whoa, hold on, time out. It's just a sandbox sound. Right. And so that, you know, so the discussion with our big insights product is it's not just the Hadoop, it's got the Hadoop technology in it, but it builds then the governance and the management and the security stuff into it. Well, the real world, the big leagues. Exactly, exactly. And that's why when we do things like we put, you know, XML store and the RDF graph store in DB2 and the time series and spatial managements and informics, when you put that new capabilities in existing systems that have all the reliability, security and whatnot on top of it, you get that immediate benefit. And so that's why you see us in some cases where it makes sense, blending those technologies right in, so you get that immediate advantage. And then you have other cases like in the Hadoop case, that's really a separate system and we've got to take that base and build the security and the reliability and the enterprise class closet service on top of it. What's your angle and the best way to do that? So for instance, let's take security, sort of narrowly. And there have been some initiatives like the Accumulo project, you know, the big table mimics out there, but specifically people getting into things like cell level security out of the NSA, there have been some projects spawned. Is that the right direction in your view or is it more trying to get data back into traditional structures? What are your thoughts on that? I'm not a security expert, but I think the trying to get data back into a box is not the right way to go. They're here. You know, it's never going to happen. Once you get this exposed, you're going to take advantage of it. The key thing is though, to be smart about it. And to know when there are cases where, especially in the big data and using Hadoop world, you may have an area that's wide open, you want it to be the sandbox and you're just playing around and snooping around trying to figure out what's important. Now as long as you're not loading sensitive information into that pool of information, you know, that's okay. Now if you want to mix in your sensitive information, now you want to apply techniques like masking and privacy that you get with like, for instance, our infosphere optum tools that originated with worrying about a structured database and data warehouse, but now have to be brought into this bigger world. So I got to ask you, because you're a database expert and John said it's become much more interesting in the last few years. We were at Oracle Open World a few weeks ago. And years ago, Ellison made a big deal about you don't want to do multi-tenancy, it's a bad idea, and then they came up with their patchwork of multi-tenancy solutions. And of course, Oracle 12 was announced and now all of a sudden, multi-tenancy is a good thing. And you need multi-tenancy in the database, not the application layer. First of all, do you agree with that? Let me start there. Well, I agree that multi-tenancy is a valuable thing and that's why I know we've had it in DB2 since the early 90s, contrary to an assertion that the new Oracle product was the first to happen. But you know, you know how those things go. I had to get that out there. So yeah, let me just point that out. Deny it and then act like you invented it. Well, right, exactly. But it gets back to the virtualization technology enables a greater level of efficiency. I'm going to bring it back to my point about speed, simplicity, efficiency. Virtualization capability allows you to, instead of having many things that are underutilized, have fewer things that are highly utilized. Now, when people talk about virtualization technology, they go right to, you know, the VMware, the KVM. Do you guys know anything about virtualization? All they say, right, do we know anything about virtualization? Right, and the cloud, and this is all new. Well, you know, the system Z, the mainframe has been a virtualized cloud environment for decades and it's still the India's industry's lowest cost for application user platform on the planet. Because you can pack the applications and the workloads and the user throughput onto that system because it's so highly virtualized. So it's virtualized at all the different layers, storage virtualization, operating system virtualization, data virtualization. So the multi-tenancy cuts across all these different things and in distributed environments, X86 environments, even UNIX environments, you have a whole generation of folks who don't appreciate all that level of virtualization that's even possible. So when new things like this come up, oh, this is a new thing. Well, yeah, it's an old thing that's new again. But if you could bring that kind of multi-tenancy through NoSQL to the cloud, that would be, in your view, a good thing. Oh, yeah, certainly. No, the multi-tenancy to increase the virtualization and the efficiency, I'm sorry, of the underlying platform means lower costs for the provider, which then can translate to lower costs for the subscriber. And if you can deliver security on top of that, right? That's nirvana. Right, that's exactly right. And the more you can bake things together and manage them as a collection, instead of a distribution, then that enables you to build some automation in, as well as the security. So you see that as table stakes to play in the big leagues, right? In terms of when you start talking multi-tenancy. Exactly, yeah. All right. All right, excellent. So we're here at IBM IOD. We're here at the Mandalay Bay in Las Vegas with Bernie Spang, who is a database expert, head of, you are now. Yeah, I am now. He's a senior vice president involved in setting strategy. We're kicking out having a great time. It's cool to be in the database. My final question for you will end the segment on a big question with some big insight, from some big data coming from you. And that is, take us through, just, you know, from your own personal perspective, not the IBM, take your IBM hat off, share with us the database arc. How's this market going to change for the students out there watching, for the analysts that are watching, for the users that are watching? What do you see as the next five years of trajectory? Shoot the arrow forward five years, be in the geek that you are, talk about databases. What's going to happen? What do you see as scenarios, and just share some vision? You know, it's interesting. My nephew graduated a couple of years ago and now is with IBM working on data analytics. My son is just going off to college next year and is interested in business, and he's got an interest on the whole data analytics. It's an interesting time, and I think it's going to be for at least another five years if not a 10-year trajectory going forward where the data professionals are going to be seen increasingly as critical to the success of the companies, their companies, their organizations. And it needs to go beyond the, it used to be I'm the management expert, I know all the dials and switches to get it to actually run. It now needs to be how do I take advantage of, not just making that one thing work, but how do I understand all the different options available, all these different technologies? How do I understand all the different things that business is trying to do with data and connect those up? So it's a great time for the data professionals to really think in terms of being data experts, not just database experts. And I think that's going to continue. And then I think from a product point of view, the database products are going to continue to what I just said. They're going to increasingly have different workload optimized capabilities, increased automation, increased simplification, as well as never ending demand for performance and security. Databases are hot, data science is hot, up and down the stack. Big benefits from down into the geek to the business benefits and business impact. This is theCUBE, we're live at Information on Demand in Las Vegas, IBM's big show in 2012 for information management, big data. This is theCUBE, we'll be right back with our next guest after this short break.