 Live from Las Vegas, Nevada. Extracting the signal from the noise. It's theCUBE covering IBM Edge 2015, brought to you by IBM. Welcome back to SiliconANGLE TVs. Live coverage from IBM Edge 2015. I'm Stu Miniman with Wikibon. Really pleased to have with me. I've got Ross Mori, who's the general manager of Z-Systems with IBM and also Donna Dillenberger. Sorry, IBM Fellow. Thank you so much for joining us, both of you. Hello, Stu. Hey, Stu, happy to be here. All right, so Ross, you know, Dave and the team caught up with you at the Z-13 launch earlier this year. I guess start with us. What's been the client reaction? You know, what's happened between that launch in New York City and now? Well, the client reaction has been phenomenal worldwide. You probably saw some of our first quarter results, 130% year-to-year growth. That's great. To me, more importantly, you know, we spent the last five years co-creating the Z-13 with clients, and now I'm positive that this reaction is because we worked so closely with them to really meet their needs now and for the future. Anyway, the reaction's been fantastic. All right, so Donna, you know, one of the things, I guess it shouldn't have been surprising to me, but you know, you kind of listen to some of the industry conversation and it's like, you know, a mainframe and power. You know, they've been around for a while. We've got, you know, it's cloud and big data and, you know, Docker is like the hottest thing out there and it's like, well, come on. We can do, you know, Linux on Z. We can do containers on Z. And, you know, you're running the research on this. You know, how does IBM keep, you know, something that no offense is an older technology? Keep it fresh, new, and driving forward? Yeah, the good thing about the mainframe is that we take something that's new and we provide the grown-up version of it. At IBM Research, we provided the Docker container for Z and the Docker container on Z is scalable. You could put thousands of containers on Z and we're also making sure that it's secure. So the grown-up version is a version that you could have more of and that people can't break into. Yeah, I have a comment, please. So, you mentioned that you've come into the technologies old. It's what's old or it's been around as the architecture, right? That's why all those applications still run. Actually, the technology in Z13 is all absolutely state-of-the-art. I mean, that's what casters up design, more than 500 new patented technologies in there. I mean, I'll take on anybody on a discussion on modern technology and it's in Z13. Yeah, absolutely, you bring up a great point there. It's funny, you talk about branding. Most people think when they hear a brand and they're like, oh, that's been around for years. It's like, you know, the Z13, it's not my father's Cadillac. That's right. Is what we're doing. That's right. You know, one of the questions Dave wanted me to bring up is you're bringing transactional data in analytics together. Traditionally, that's been a complete no-no. So, you know, that's taboo. How are you doing that now? That's a great question. It's one of the reasons I wanted Donna here as well. It's actually really exciting. I mean, for the last two decades, right? You know, we've been encouraging people to, you know, your operational data is created on the Z through an online transaction processing system. You really care about throughput and response time, but you didn't want to do your queries, your analytics there. You moved the data off the platform. Well, we put a lot of effort and innovation into the software and the hardware over the last three years in particular to make it so that you can run your OLTP system. You're not going to affect response time. You're not going to affect throughput. You can do large queries. You can do what we call in-transaction analytics. So, a complex, scored, analytic that's part of a transaction done in real time. A lot of work went into it, a lot of innovation. And, you know, Donna and her team have been at the forefront of these breakthroughs. Yeah, at IBM Research, we're putting advanced analytics on Z, things like deep learning, things like cortical learning, things like graph data stores. So, the most advanced analytics so that it could run against enterprise data, the data that you really want to see insights on. All right, so Donna, yeah, I mean, please go ahead, yeah. And I think the real, it's really cool technology. The business value is using the insights from your data at the point of the transaction. So, while you're doing signing someone up for a new credit card, or you're doing someone submitting a medical claim, or you've got a transaction going on and you want to do cross-sell, that's the point where you have the customer there and that's where we can really make a difference. So, I'm wondering if you could help unpack for us a little. Something I've been looking at the last few years is, we have the great wave of virtualization that came on the X86. There's certain application that stayed physical. Now, there's some new applications that many of them are physical, some of them are looking at containers, sometimes it's containers on top of virtualization. How do you look at the mainframe in what application portfolio, things like analytics, how does that fit in architecturally? Analytics is a broad spectrum. So, when you have to have the analytics change the data, then you want to make sure that data persists, that it's transactional. So, for example, the new thing now is to read data from flat files, but as computer scientists, we all know that's the worst way to read data because that's just going to incur a lot of IO. So, that's why Hadoop is going out of favor, Apache Spark is coming in to read all the data in memory. When you start reading the data in memory, that's when you really need it to be in containers, to be virtualized, but then what grownups are going to realize about Apache Spark and in-memory data is that once the node fails, that data and those analytics are going to be gone. So, then the next thing is to be able to persist that data automatically. And again, that's where the mainframe comes. We already have an in-memory analytics server. You could save the data transparently, we'll cache the data, we'll cache the most frequent queries in analytics. So, your use of analytics is going to cause the mainframe to change how the data is transparently persisted for you. Yeah, if I can just poke at that a little bit, because it's interesting. We sometimes have these pithy little statements we talk about IT. And if you talk about distributed systems, you talk about things like Netflix. Hardware eventually will fail and software will eventually work. And we expect pieces to fail, but I've talked to a lot of Z customers this week and they're just like, we don't have downtime. That's kind of the traditional storage world is failures are no good, but how do you reconcile that kind of distributed architecture? We need global environments with what you're doing. In the distributed architecture world, they assume that the hardware is going to fail. So, that puts the onus, the burden on the application developer to make sure that whatever state they save is going to be sent over to the next version of the software. In Z, we don't put that onus on the application owner. We have these built-in mechanisms to transparently persist the data and load balance the data. So, would you rather spend like 30 or 40% of your code for availability or just focusing on application logic? Yeah. I mean, interesting, actually I got a note from the crowd said that Sysplex was the first large system clustering technology when it was introduced. You know, how are you positioning that today? Parallel Sysplex is alive and well. We're providing failover technologies for our Linux on Z systems, which there's nothing that compares with the parallel Sysplex on Z. So, we're providing that to the Linux world as well. Yeah, I think the important thing that's grown out of Sysplex, especially in the regulated environment of banking today, where they're required to have uptime and backup data centers and all that, extending Sysplex now over geographically dispersed areas, having active, active hotspites. That is really, really prevalent. We have more than 750 full customers running fully geographically dispersed mainframe Sysplexes. So, it actually is kind of the industry standard, if you would, if you really want Uber high availability, super high availability, especially in a regulated environment. All right, so Ross, how does mobile fit into this whole discussion? That's a great discussion. So, you know, we've got the transaction systems on these mainframes, right? And everyone has their, every business wants to mobilize their applications, whether it's for their employees and their enterprise, or it's to reach their consumers, their customers. The tie there, though, is that usually, if you want a cool app, you know, you're going to write a system of engagement, it's going to have a beautiful UI, it's probably going to run out of the cloud. But if you want to do something real with my business, you want to check inventory, you want to check your account balance, you want to transfer money, you want to do a transaction, it's got to go back to the transactional system. So, we've built connectors in to make it very easy for mobile app developers, whether they're on-prem or in the cloud, to connect to these backend systems and do it securely, because everybody cares about the security of their personal and their business data. So, secure connection, easy to connect in, and the backend systems are built to scale to any spike that mobile may drive. Okay. And at Research 4 Mobile, we've helped build the deterministic random number generator in the Z hardware. And what that provides is an unlimited number of random numbers so that every query in your session could be encrypted. So, even if they get the key for the first query, the second query is going to have a different key. So, Z is a platform where you could have an unlimited number of these random number generators. Okay. So, next thing I want to discuss a little bit is the R&D cycle. So, mainframe R&D traditionally takes a long time. We've been trying to shorten release cycles. How does that fit into the conversation? Well, we do a lot of things in parallel now. And of course, from a hardware point of view, the old days with five or six passes of hardware and having to do the test that way, we do a lot in simulation and emulation. So, we're really doing two-pass designs. But we really have to have to throttle how often we bring out a new system because can our customers, how often can they absorb a new one? So, we're on about a two-and-a-half-year cycle now and our customers tell us that's about the right timing for them. And so, that's what we'll probably be doing for the future. Yeah. And we also have beta development sites where we put out code for customers to try on Z. And they could download that at will. And we put out new things for them to just play with multiple times a year. Okay, great. So, at the show here, are you, the customers I'm sure are asking you, what's next? What are you working on? What are the cool things we should be looking for in the future? What are you telling them? We're working on enabling the mainframe to try to anticipate what patterns you have so that it could help you find new patterns for your customers, new insights in their behavior. And also, personally, looking to see how you use systems, not just to extract information, but to also help people become more healthy and help people be more compassionate as well. And absolutely, and what I tell people right now is we continue to drive open technologies, expand our Linux ecosystem. We have KVM coming as a third virtualization environment for the mainframe, working with OpenStack to make sure that we can be managed properly and fit into the hierarchy. So we're going to continue to drive our Linux environment and our open environment more and more and more because it's now 28% of the installed capacity around the world of mainframes is running Linux. So it's a really important open environment today for our customers and I really see a lot of growth there. Yeah, we also have natural language interfaces to the mainframe so people could just type in English questions and then the mainframe will be able to provide the results just by you don't have to be an SQL programmer or a deal one programmer. You could just talk to it. Wow. Okay. I don't even know what to say there. I don't see that coming out of windows. It's a friendly mainframe, we call it the mindframe. So Ross, I guess one of the things that comes back to our early part of the discussion is the perception of the mainframe. How are you shifting that mainframe to the happy mainframe? I didn't see a dancing mainframe here at the show but how do people understand kind of the mainframe of today and tomorrow? Well, I think what we're trying to do is well we show that beautifully designed black box but I think more importantly we're trying to position the mainframe as part of the mobile app economy because it is. We're trying to make sure that people know that the mainframes advanced analytics are superior and can really bring great business value and from a cloud environment. I mean talk about security and scale and virtualization for a private cloud, a public cloud or a hybrid cloud. We're trying to make sure that we're seen in the right light for all these key trends and so I think that's the most important thing that the technology's brand new but most importantly the trends that are driving our customers' businesses were a healthy key part of that growth. Yeah, and you think about it, right? We have the fastest microprocessor in the industry. We have 10 terabytes of main memory. I do collaborations with other researchers around the world on using the mainframe for global climate modeling, for genetic sequencing and if you think about all those types of calculations in the past would take days and now they only take minutes. You could crunch all this data in real time so that you could have insights that allows you to do things that you couldn't do before. We did some things in Z13 that probably nobody expected. There's a full vector instruction set in there, full SIMD instruction set that was really put in for computationally intensive computing for analytics so we keep putting more architecture and more hardware behind the mainframe to make sure that things really, really are fast especially when it comes to analytics. Yeah, some of the global climate modeling people I've worked with when I just gave them a Linux virtual machine on my mainframe and they asked me what system is this running on and I said it's a mainframe and they said it runs faster than a cray. I mean that's how powerful it is. Yeah, so Donna, I mean IBM is so well known for innovation. I mean big celebration about 100 years recently, big thick book on it. I tell you on the plane right out I've been reading Walter Isaacson's The Innovators which talks about really the history of the computer, talks about how IBM created so much of modern computing. What's the role of IBM research today in really helping the really global innovation in IT? To transform our systems to not just compute but to provide insights. It's not something that just does calculations. It's something that helps you live a more healthy life and live a more fruitful life, help our businesses prosper as well. All right, so Ross, I want to just give you the final word, we're getting the hook. People come away from IBM Edge 2015. What should they think about Z? Well, I hope they see that we're very much alive, very healthy, important part of the global economy. Again, and not just the transaction processing hub of the world that we are, but extending out into the mobile app economy with all of our cloud dynamics and of course with advanced analytics. So just want to make sure that they see how relevant Z is now. More relevant than probably any time and it's more than 50 year history. All right, Ross and Donna, thank you so much for your time. Wish we had more, hope we can catch up with you in the future. Really appreciate all your customers that have come. Talked about really global deployments, mission critical applications and a lot of modern things happening on the Z. So lots more coverage coming here from IBM Edge 2015. We'll be right back after this quick break.