 Live from Las Vegas, Nevada, it's theCUBE at IBM Interconnect 2015. Brought to you by headline sponsor IBM. Welcome back to Las Vegas everybody. This is theCUBE and we're live at IBM Interconnect we're at the Mandalay Bay. I'm really pleased to have Dion Newman who's the Vice President of Worldwide Z Marketing and also Ron Perry who's the CEO of Radix International gentlemen. Welcome to theCUBE, it seems like just yesterday we were in New York City for the Z announcement. You know it's great to be here, thanks for having us. So that was a great event. We did it at Jazz at Lincoln Center. I'd never actually been inside, it was a phenomenal venue. You guys really know how to throw an excellent party, great announcement, customers were pumped up. So you must have felt really good after the results came in on that show. You know more than that, you know the reaction to the Z13 have been off the charts. You know we had probably 600 articles around the world written on in a 24 hour period. Very very positive response. I think really a repositioning of what the platform's all about. Yeah so and Ron we met actually at dinner before and we want to get into sort of the story but from a customer perspective, what do you get out of an event like that? Well I think we learn things certainly but beyond that it helps us, it really reinforces the reasons we made the decision. You know it was really just about three and a half, four months ago when we made the decision. We took the information we had, we went forward with it and then to see all of the other implications. Very positive, very pleased. You know mainframe cycles, the lot goes into it. I think you had hundreds of patents. Yeah, about 500, this is the last year. 500 patents, just as part of this announcement, it's mind boggling, a huge percentage of the customers running Linux. So the theme on open, so what is this announcement, this Z13 announcement, what's it mean for IBM's customers and IBM as a company? Well I think, let's sort of start at the beginning. This is a totally new system designed from the customer side. We spent about a billion dollars here developing this new generation system and you know it's really been over about five years. You know these cycles, we're in development of not just next generation but the one after that. We work with probably 60 clients in our leadership council who help prioritize where we invest and what we build. And it became pretty clear to us sort of two and a half, three years ago that there was this change going on driven by mobile technology. That mobile and the internet of things were going to drive massive transactional growth. And so this system is really built for that mobile scenario. But the other thing which is different here and I think is the thing that's really resonating for clients is for the first time it's practical to do your analytics where you're doing your online transaction processing. We're bringing that together. And you know that's a game changer. You know that really allows you to get to real time analytics. It allows you to deliver better personalization in those mobile apps, better insight. And I think that's one of the things that's really exciting clients. Let me hit the other thought here which is you mentioned you know Linux and you know the take up of Linux. You know on this news Z13, we can run 8,000 VMs on a single frame. You know we just had a client, CCOV, the largest credit union in Brazil, 2.5 million clients. And they're saving $1.5 million a year just on the energy bill. Little on the software savings that they're getting through consolidation of their service up. They had about 400 servers. They've consolidated all that up into one Z system. The savings are phenomenal. So Z's alive and well. I want to dig into Ron's story now a little bit. I look at you as largely a data company. So you talk about analytics and transactions coming together. Talk a little bit about Radex. Let's remind our audience what you guys do. Sure. Radex is a company that provides a software as a service to airlines. And that service really is airline reservation systems. So it's the core of the airline's distribution. And what we're doing is selling our services to airlines. So we're not a travel agent. We're not an intermediary for travel agents. We're providing the core systems to the airlines. We have about 40 of them all over the world on five continents. The vast majority are in the developing world. And if someone is flying one of our airlines if they make a booking at their travel agent or they go online to make a booking or they check in with that airline, it all comes back to our computers in Orlando, Florida. Everything's handled there. So what I didn't realize before we spoke that night is how arcane the system is. When you get a ticket and you ticket it on a different carrier, the data and information flow is so not real time. And the ability to monetize that information accurately and provide ancillary services around it, it really didn't exist before Radex, right? That's what we're doing. Yeah, pretty much. We implemented the first graphical interface, internet based system back in the mid 90s, 95. And you really have in the industry two types of airlines right now. You have one type that uses a traditional business model. They have e-ticketing. They basically are dealing with paper tickets with their virtual so you don't see them. But the flow is the same. It's a very strange flow because a paper ticket had to be in one place at a time so we have an engine that simulates a paper ticket. It's like in the old days when people would just take their brochure and make a website out of it. Absolutely. We put our brochure up on the website. Then you have another type of airline that is purely ticketless. It's just an electronic transaction. It's just like booking a hotel or booking a car. We're somewhat unique in that we support both. And we support both for one airline. So certain things they may want to distribute in a ticketless fashion. Others using a traditional model. Now when you started Radex, how long ago? We started in 1993. You know a lot about mainframes. Yeah, my background was mainframe. Long history mainframes. And you chose to start on different infrastructure. Non-difference, so-called open infrastructure. Exactly. But if I understand it, you've swept the floor of that infrastructure now and are moving, have moved to the environment. Well, if you were to look at the early 90s, what you would have seen was communications was expensive. You had relatively little bandwidth available. And the degree to which someone was moving from mainframes to servers, basically we were looking at local area mainframes replaced with local area servers, local area networks. And so early to mid 90s, we were still kind of in that frame. And software on the mainframes were very expensive. It was a very different world. And so we chose to develop something internet-based, Windows front end at that point and really attempting to build in a very different way. What we've seen though is, building out with servers, you wind up getting to a point where the benefits begin to diminish because as you add more and more servers, the complexity increases. And you wind up with all of this multiplicity of multi-vendor components, constantly getting changes being made to them by their manufacturers. And then you're trying to manage that as if it's a single entity, very, very hard. Now you have a lot of custom code that you've developed. Oh yeah, we have many millions of custom code. And you've had to import that over to a Z environment. And so I think back to my days of consulting with large mainframe customers, and one of the challenges that they always had was that if they had a lot of COBOL code, they couldn't, they had to freeze the code, but they didn't want to freeze the code. If they didn't freeze the code, they were in big trouble. All right, because by the time they migrated, so there was this endless loop. And so they said, all right, forget it. It's too complicated. You didn't run into those problems. No, actually that was one of the most amazing things. Going from mainframes to servers was an 18 month to two year project, rewriting code and the rest. Now we're taking, we're running on Linux on the servers and we're running on the database, running on Linux. And so we're just loading in more servers. They just happen to be inside of a Z box. And so instead of provisioning all those individual servers with all the wires, the routers and everything else, we're just duplicating or replicating the database and turning on the standby. So this strikes to the heart of the Z strategy that was initiated many years ago. Talk about that and why it's so important. So 14 years ago we put Linux on the mainframe. It was a side project in Birgling home. They called us up and said, hey, we've put Linux on the mainframe. We said, why did you do that? Good idea. So fast forward now, there's over 30 million MIPS that are out there from the capacity standpoint for mainframes. We're at a point where 27% eight million MIPS now are Linux. 27% of all mainframe capacity is now Linux. Now that Linux is not a special mainframe Linux. It's Linux, it is, Linux is Linux. And by the way, Java is Java. There's another six million MIPS, some 20, 22% of all that mainframe capacity, which is Java. So we actually have the system that basically 50% of all mainframe workload now is either Java or Linux based. And that's bringing all sorts of workloads onto the mainframe. We see a lot of clients consolidating, lifting up big server farms and just pulling them in and dropping them on a mainframe. And the balance of those workloads are stuff that works. You don't want to mess with it. It's secure, it's fast, I'm not going to touch that. So you're seeing new growth as a result of Linux, Java, modernization of the platform? Yeah, we're seeing in excess about a 30% caga on Linux growth on the platform. We see big growth there. And the arguments are several. The first is you get the reliability of the Z system. The second is huge software savings. We're seeing consolidation down. We're able to, probably on every IFL, we're able to bring across 40, 50 VMs on a single integrated facility for Linux. The maintenance becomes much easier. You don't have people running around with fixed packs and sort of fixing things that break and plugging UIs in and URouters in and connecting things up. It all happens and all of the management can be done within that single frame. Each space savings, big energy savings. So we're wrong, okay. So people say, well, aren't mainframes expensive? What are you finding? Well, actually, we approach things basically on a lease basis, typically. And we look at a three-year horizon. And if you look in that three years and say, okay, we're going to buy this one box and we're going to spread out the cost over three years. And then you figure there'll be some additions during that time, but all inside the box. And you compare that to a server farm where you're constantly adding servers and you're constantly dealing with the changes that come. And the servers that you add at the beginning may be from the same vendor, but they may be very different two years later. And so you're dealing not just with the multiples of your vendors, but you're at different versions and the rest of these servers, different internal drivers and the rest. You start adding that up, the expense is significantly greater. Well, something that's space of the industry, the part that you sweep in the foes, they're all sort of building what's so-called converged systems trying to build mainframes. Right, essentially. Exactly right. That's do-it-yourself mainframes, no question. Yeah. So I mean, what do you make of that trend, right? You see that coming in, you say, okay, that's validation, but you've got to compete with that. Well, let's, I mean, Ron, if you look at, I went to ask you to talk financials, but if you were going to estimate your savings over that three-year period, like what would be your estimate? In terms of actual dollars? No, in terms of percentage. In terms of percentage, it's 35%. 35%. 35% reduction in total cost of ownership between what we had projected with the server farms and what we, at this point, are seeing with the Z. And some of the things were surprises. So for example, labor cost, it is so much easier to manage a Z than it is to manage a complex server form. And I mean, it's like fighting brush fires, getting reliability out of a server form. You're constantly dealing with little different things happening that you've got to face. Yeah, actually, anytime we've done a lot of TCO work in our day, we all have, obviously, looked at it. The big nut is always labor. It's a very labor-intensive business, the IT organizations. So it comes from labor savings. That's one. How do you quantify that? Is that a redeployment of labor? Do you not hire new people as a result of that? Do you actually reduce FTEs? How do you drop that to the bottom line? Well, we're not going to reduce FTEs, but there are hires we're not going to make, and that's fairly easy for us to project because we knew what our plan was. We are redeploying some of our folks to areas that we wanted to address that we couldn't, and that now we'll be able to. But labor really is just one piece of it. Where else? Well, power. So the total power of the Z right now, power consumption, is roughly a third of what the servers are taking that it's replacing. Okay, and we have our own data center at our headquarters. We also use space at a co-location center. The co-location center charges for power. So that's a significant factor when you're looking at it. And then space, you know, I mean, as we grow, we had projected taking significantly more space. Now the space is inside the box. Well, what's your take on cloud? Well, I've talked to a lot of CEOs, and they said, I just want to outsource the whole darn thing. Well, I mean, I think that the logical point of delineation is the API. So from the API back, API database and so on data warehouse and all of our messaging, all of our payment processing communications and so on, that'll be the main frame. Anything in front of that, what I'd be able to refer to as the systems of engagement in the cloud. Now we've already dabbled a little bit with the cloud, but ultimately our plan is to move everything on the front end to the cloud. For us, it's very practical. We have airlines all over the world. If we can move the hosting of those systems and engagement close to those airlines, we can reduce the latency. Reduce the latency, you shorten the transaction time. If you shorten the transaction time, you reduce the load on the systems of records. So in thinking about this nice 35% TCO savings, but if I compare that to what you just described as the business value, I would think the business value would deliver telephone numbers relative to the cost savings. Is that a fair assessment? I'm not sure I quite understand. The value that you're getting out of the reduced latency and the productivity of the business and the dollar value that that delivers is, I would think, much greater. It's pretty substantial. Then the cost savings. Yeah, no question. I mean, when you start looking at increasing the performance, so much of what we see now is somebody with a cell phone who is trying to find out, am I on board that flight? Did I get my seat upgrade? Checking in and doing any number of other things. But particularly when you get to the buy side where they're making the initial purchase, that's the difference between getting a booking or not. And since we charge a transaction fee, we're really just a large virtual airline that has 40 fulfillment partners all over the world. We want them to grow. We want them to get as many bookings as they can possibly get. And if the system is slow, there are fewer bookings. So the revenue side, that's not even taken into that TCO discussion. What more can you tell us about? Well, let me actually start with, yeah, mobile's a big trust. IBM is a company, but specifically the Z Group. Talk a little bit about mobile. I want to understand, Ron, how it's affected your business, but what's the mobile trust? Well, at the end of the day, the new battleground is client experience. It's about being available always, always on. 24, 7, 365, when I reach my phone, when I get out of bed and check my bank balance, or I decide that I want to buy something, if the site's not available, I'm going to go somewhere else. So there's a number of things that we see. You need speed. That's probably the first thing. We expect that in one second or less that we'll get the response we're looking for. So when we go to that app, we hit the button, we want speed. Secondly, availability, that 24, 7 that I was talking about. Then you start talking about differentiation as you deal with me. So the demographic of one. So think about personalization. And that's where this analytic story sort of comes in and why we brought the analytics together with the transaction speed, because you want to personalize the information. You treat me as a single customer who's got a history with you, who's got particular likes or dislikes in the travel industry. Maybe I like hotels downtown. Maybe I like to sit in this sort of a seat. Maybe I like to go to the theater. I have all of these preferences. Treat me as a customer. You don't treat me as some random person who got on my website. So personalization is really important. And then security. We're putting all this information, our credit card information. Maybe there's a social element, your social security number element to this thing. So security matters here as well. So one more I want to give you is scalability. So I heard this great story about this bank in Australia and their marketing department deployed this app. And I'll tell you a little bit about it in a second. But the backend guys, the guys in the glass house, they suddenly saw spikes in the system on Sunday night. And they could not work out what was going on. Now they're running a Z which automatically just, move more resource and manage the spike. But it turned out that they'd run a TV ad that showed that if you shook your phone, you didn't have to enter the app. It just brought your credit, your banking balance up onto the front of the phone. And so all of these people watching TV, watching this high rating, whatever the show was, were sitting there shaking their phone and suddenly there were tens of thousands more transactions running through, doing account balance checks. And so you need to get a cope with those spikes. Because the last thing we want is denial. I can't do it right now, systems breaking. You just can't have that. So what we focus on is delivering that performance. So that's feed. We can do 30,000 transactions per second or two and a half billion a day. That's like 100 cyber Mondays. You can do on one of these Z13s. We're focused on availability. That's always been a hallmark of our platform. The best security in the industry, which is again a hallmark of Z. And then bringing this analytics in, it's about personalization. So we think we've designed something for the modern sort of digital economy and we think it's going to be very successful. Yeah, you're hitting a lot of those trends. I want to come back to the mobile piece and talk in the context of your industry. Your customers, they've gone through some tough times in the past decade, right? With fuel prices going through the roof. But now they've seemed to rationalize their cost structures. Fuel prices have dropped dramatically and you're seeing the airline stocks are booming. Things are good. Now you have also mobile as a tailwind. The economy's obviously doing better than it was five years ago. What does that mean for your customer's business? And obviously ultimately yours. Okay, what mobile means is massive additional demand on the systems of record. Massive. And I'll give you a statistic. Okay, airlines talk about look to book ratio. So if you were to go back when bookings were made through the call center, rule of thumb was three to four calls for one booking. So look to book ratio of let's say four. Today, two things are driving a massive look to book ratio. One is mobile, the other is the online travel agency. And the combination of those two on a typical day, we will see 200 or 300 looks to one book. On a sale day, it'll get to 2,000, sometimes more. We had one airline a few years back that gave away a million free seats in India. They had 70 million availability requests, which is a look. They got 85,000 bookings that day. So it's killing the normal ratios in your business. No question. But it's driving business through the room. And it's totally unpredictable because the difference is with mobile it's unconstrained. When somebody had to go to a computer and sit down somewhere, that limited them in times of access. Now they're standing in a bus line, they're pretty much anywhere, and they're using our systems and spawning hundreds and thousands of transactions. This is a great story. So it just really underscores the tension in your business, how you can put in an infrastructure that does matter. That's a big theme of IBM's at the show. Infrastructure matters. I'm an infrastructure guy, so I always help get matters. But so, it really matters when it doesn't work. It's like that. So people notice, but your business would be quite a bit different without this transition. You'd have to pour a lot more, I'm hearing it right, a lot more people. You'd have to deal with a lot more security issues. You'd have to spend a lot more money. And it's taking away from an opportunity. That would be taking away from an opportunity that you see as a CEO. That's what you care about. Yeah, no question. Plus it creates a competitive advantage for us. People just intuitively know the mainframe is more reliable. They know it's more scalable without looking at any of the numbers. And so it's positioned us very differently to have modern software running on a rock solid base. Well, Ron, it's a great case study, bringing that analytic piece into your business with the transactions, the volume. Dion, I'll give you the last word. What's the big themes that you want people to remember? I mean, you touched on a lot of them. We're here at Interconnect. We've got all kinds of action going on in cloud and mobile and analytics. So what's the final message? Well, you know, so I've talked a lot about mobile and I've talked a lot about analytics and we talked about Linux consolidations and a little bit about cloud. But there are a couple of other things which are worth mentioning, which are new news. We've done a lot around software pricing, you know, really game-change stuff. So we've brought out what we call iCAP or it's basically containerized pricing for incremental workloads that come onto the platform, essentially isolating that workload, making sure that the costs associated with that software are independent and we're getting a lot of excitement from clients on that. New flexibility like we've never had before, multi-flex pricing that allows, you know, in-country to consolidate MIPS up for a mainframe client. This has been a big ask and so we announced that as well. Bluemix for the first time, full Z participant participation now with Bluemix. So that's another new thing we've added. Another key pricing action we've taken is around memory. So in this new system, we've got 10 terabytes of memory as a maximum on the system before the maximum is three terabytes. For many of our clients, that'll mean they can go three X to five X what they're currently running on their systems. So we really want to encourage that. You know, as we see more Linux workloads going on, we've added SMT to the platform, we think there's going to be a lot more Linux coming onto the platform. It'll help with Java performance as well. We've taken the axe to memory pricing and, you know, for many clients, it'll be 30% of what they paid for memory in the past. So you could basically, what you used to buy your memory for, you could now get three X. You know, in fact, if you get more than that, you'll see even bigger saving. But, you know, for many of our clients, there'll be dramatic savings around memory. And, you know, that's going to deliver immediate performance improvements. It'll mean less, you know, less pressure on the processes. You know, it'll mean we're spending less time moving data around. And, you know, we're seeing very large projections on performance improvement because of that additional memory in the systems. So mainframe cycles are obviously a big deal for the Z division, big deal for IBM. I talked to a number of customers at your event. And several said, yes, I'm buying, not I'm thinking about it. I'm buying, in fact, I'm buying four site unseen. So that's great. It's great news for IBM. It's great news for the Z platform. And I think great news for customers. So gentlemen, thanks very much for coming to theCUBE. It was really a pleasure having you. Great story. Thank you very much. All right, keep right there, everybody. We'll be back. This is theCUBE. We're live from IBM Interconnect at the Mandalay Bay. This is theCUBE.