 Live from San Francisco, it's theCUBE. Covering IBM Think 2019, brought to you by IBM. Okay, welcome back everyone. We're live here in San Francisco for theCUBE's exclusive coverage of IBM Think 2019. Jeff, our student men and men, our next two guests are Greg Lodko, Senior Vice President and General Manager of the Main Frame Division of Broadcom, formerly with CEA acquired and Clayton Donnelly, Head of Security and Integration at Broadcom, both formerly of CEA, big acquisition, big value, you guys welcome to theCUBE, good to see you. Thanks a lot for having us. So we just talked before we came on camera here at IBM Think, a lot of content around software, AI, cloud, systems, and software working together. Kind of the thesis of the Broadcom CEA acquisition, that move was a big one. A lot of analysts liked it. Your thoughts, how that's playing out here in San Francisco. Yeah, I think it was really interesting. If you look at what Broadcom has gone after in the marketplace, is they're not looking for the flash in the pan or trying to chase the next new thing. They're looking for core businesses or components, software, products that they believe have real staying power and will be around for a decade or more into the future. And then they want to invest in those and nurture. They really want to be, even if it's in a niche space, they want to invest in something where they'll be number one or two in the marketplace. It's interesting, you're looking at the market with cloud, you see scale, you see data, all this stuff we've been talking about for years and years. But it really comes back down to systems and software working together. Cloud's one big complex distributed system. Software abstractions can maybe extract away those complexities. But in a day, it's software plus large scale systems. This is kind of what's playing out. This is in the wheelhouse of what you guys are doing. Can you guys add some color to that trend and what needs to happen to make it more and more viable, more performant, easier to use? Yeah, I mean, I think that what we see is that customers aren't having a lot of problems with individual pieces of software. They're having problems when they put all of this together. So if you look at even to your question about rock on a moment ago, they sort of came in enterprise software through the data center. Because they're providing everything from AI chips to fiber channel networks and other kinds of things that are running people's networks at the very largest scale. And what they realized is when you get to the enterprise software level, the customers have such challenges because they don't get to cherry pick just the cool things or the easy things to go integrate. They've got everything from mainframe to client server to interior to whatever they've picked up over the years. And that stuff has to work together seamlessly to get the kind of value that makes sense. And that's why I think when you start looking at kind of our focus is on helping customers bring that together to get that value. Great, talk about the hybrid IT environment because what we're getting at here is I got to make the legacy work with the new. But the beautiful thing about cloud native and some of these new microservices and containers is you don't have to kill the old to bring in the new. There's a great abstraction around software now that's making them work together. But yet the new stuff can work really great. This is the kind of the new architecture. Your thoughts on this? Yeah, I mean, obviously you're here a lot about that here at Think, right? They're talking all about hybrid IT or multi cloud. I think there's a stat out there that, you know, 75% of the large enterprises in the world say they'll be have multi cloud or hybrid IT environments by 2020. I think they all have it today, right? You think mobile to mainframe, right? There's not workloads that work in isolation. You pick up your phone and you go to check your balance. It's going to kick off a transaction that's going to go to an edge device or an edge server. That's going to go through a network and maybe hit another server. And eventually it's going to go back to a mainframe to check the balance or to transfer funds or something like that. So they're having to deal with it already today. And there's two kind of sides of the coin. You want that interaction for the developers to be common across those platforms. Yet you want them to be able to leverage on the power, the strength, the security of the underlying platform without having to know all the gory details, which is why it makes a lot of sense for us mainframe and distributed if you look across what is the CA Technologies portfolio that Broadcom acquired, a lot of the capabilities that we have are the same capabilities that work across those environments so that the enterprise customers can interact with it one way. Yeah, Clayton, when I hear this environment there's certain things that I need to worry about everywhere. It's my data, how do I protect my data? And of course security is one of those areas where there are lots of different environments and unfortunately there's lots of different considerations depending on which clouds I have, which environments, mainframe, x86 power, all have different considerations. The mantra I've heard that seems to resonate is security is everyone's responsibility, up and down the stack from the chip level all the way through the application. So explain where CA Now Broadcom fits into this picture and lives in this even more heterogeneous world. And by the way, I totally agree. The multi-cloud is what customers have today. I mean, if you look at it, I mean, customers are building out, say new mobile applications and building them as services in the cloud and so forth. But what we're finding is that the transactions and other kinds of things are still happening in some of these other environments. Maybe those environments still live in a data center, maybe they've been moved to like a private cloud, maybe they're in a public cloud, running on IaaS or some other kind of bad as a service. What we're finding is that each, that transaction has to be protected. The API that gives you the ability to call that transaction from a mobile app needs to be protected. All of these things need to be protected. But then you need to be able to orchestrate that, make sure that you're laying down those bits, protecting those bits the same way every time, testing them the same way every time. And I think that if you look at what we're looking at and our value is really in digital, infrastructure management, right? Because you're bringing all these pieces in, cloud, multi-cloud, mainframe, all of these environments that you have, you've got to have a way to operate it, manage it as well as for security. Yeah, so Greg, when I look back my career, there's something that's been repeating a lot. It's like go back to the 90s, it was like, okay, what was some of the reasons why the XSP's failed? It was like, well, it was networking security. When cloud happened, it was, well, security and management, how do I figure out some of these? Management of a heterogeneous environment has typically been a downfall in IT. It's something that we've struggled at as an industry. So why will now be different? How is the industry helping to solve that issue and simple is something that we keep trying to hear but actually achieving it is pretty challenging. I think it's fundamentally realizing that the core large enterprises in the world today are using mainframes, right? And some of them have tried to migrate some things off and it's not about complexity of migrating it off, it's about whether or not you can land somewhere that has that same security throughput, resiliency, all that kind of stuff. But if you recognize that you're going to have these systems interacting and you recognize that we have to make it easier for people, whether they're coming out of university or they're coming from a background of distributed or open source, you want to make it easier to interact. It's what's informing everything we do in our strategy and mainframe. So we talk about open, frictionless and optimized. So it's all about the idea of that mainframe system and those processes that we're running, whether it's DevOps, whether it's databases and the tools, whatever we're doing, the security, the analytics that we're doing, that has to be open and be able to interact with other people's tools as well as other people's platforms. Frictionless is all about the idea of you got to make it easy to do that interaction. Somebody that comes at this from a non-mainframe context that maybe knows, I call them the cartoon characters of open source, Git or Gulp or Jenkins or whatever, that they can use that to interact with the mainframe and leverage it. And then you want to be optimized. You want to make it for the real deep technical professional to get the most out of it and focus where the expertise is or for the novice to not really have to need training wheels but to be able to ride that bike right away and perform the thing. So all of these things, you can see how they're kind of informed in setting that tone of thinking about a hybrid IT environment and connecting that mainframe in across, not sitting as an island onto itself. I mean, you bring up a good point, a couple points. One is distributed computing. It's been around for a while, mainframes. I mean, I'm old enough to remember that I was part of the client's server way. We used the point of the mainframe guy, ah, you're going to be dead soon. Most of them kind of went away, but it never died, right? We all know, but there's a renaissance. The rumors of my death are greatly exaggerated. A lot of them did go down a little bit, but they never really died. But here's the thing. There's a renaissance in mainframe because of cloud computing and cloud operations. If everything's cloud operationalized, then essentially you have a big, one big distributed computer called resource and edges that are subsystems. So the notion of buying a mainframe isn't a platform decision. It's a the right tool for the right job kind of decision. People are not looking at mainframes as a bad decision if it fits. So it's not like everyone should buy mainframes, but if you need it, you need horsepower. So the question is, begs the question is, why is there a renaissance in mainframes? What's the reason why people are buying them? Is it because it fits into a certain position? Is it certain scale? Is it because it can plug right into the cloud and be a big resource? I think there's also a realization. Think about if you're the newer CIO and you start looking at your estate and you realize that this mainframe thing that you're spending 20% of your budget on is actually doing 70% of your processing. That you kind of look at it and you go, wait a minute, that's really cost effective. So then you start looking at, well where is it most cost effective and does it make sense to use it there? And then when you can tie it into everything else, when you can get the same types of security tools and lock it down and lock the interaction down, you say, hey, this might, but this might make sense for me to do it. And I think it just ends up being dollars in cents. And then the resiliency, right? I mean, when people aren't having that downtime. Clayton, you've got to run your business. You want uptime if you're in e-commerce. You want high secure systems. So it really is the right tool for the, well, like thing for the right job. Is this happening? Give us the update on, are people buying more ratio because it just fits better? I think the part of it also is, why fix what isn't broken, right? The main friends running there, it's up, it's providing transactions. I think what you used to have is you used to have this impediment to getting access to it. You need to find some old Kobo guy, you need to find all this other stuff because you had your business process. Kobo programmer but now it runs Linux. It's like a foreign language to some people, right? You say Kobo, it's like I have to learn Chinese. So what we've done though is we've made it so you don't have to learn Kobo. You don't have to learn some specialized thing. You can come in with APIs. You can come in with the technology if they teach kids in elementary school to use JavaScript and other kinds of things to come in and access those same things that are now in the main frame. It's basically a big iron and the old expression. Big horsepower. It is. High horsepower, high throughput, high resiliency. So Greg, I heard you talk about things like DevOps that you fit in this environment, absolutely. We've tracked it. I remember when Z-Linux on mainframe rolled out 15 years ago, you want to do the cool new Docker things, absolutely. But if I look at the DevOps people, the people that are going to pay for this, a lot of times they say, well, I'm used to more of the cloud model. How do I get, you know, they move to an OPEX model. We're still early in that trend, but does Z-Series mainframe, will it fit into the new modern paradigm from a CFO standpoint? I definitely think so. If you look at a lot of the stuff that's going on in the marketplace and even concepts that we're testing with clients today around, you know, you can refer to it as consumption-based pricing or value-based pricing, you know, looking at how much you're actually using and then charging for that with a known, you know, hey, if I grow my capacity this much, how much am I going to pay? Or if I go down, am I going to be able to redeploy those dollars elsewhere? All those constructs are stuff that we're already working with customers today on. So it is very much the idea of a cloud-like environment that can either be delivered on-prem through you buying your own hardware, or, you know, there's IBM that has the Z-Cloud, there's folks like Ensono, et cetera, that have clouds that have mainframe up in them today. And the developer environment clearly is going towards infrastructure as code, which is the abstraction of ways, just programmable infrastructure. They don't care where it's almost fast, right? Does it matter? Does it really matter? Look, we contributed to Zoey bright side, right? That was the command line interface and everybody was like, oh my God, you know, they thought maybe we had some executives that were sitting back that had this brilliant idea. We were actually using agile methodologies in our development and we gave, in each programming increment, we gave the engineers time to do what they wanted to do. You know, one sprint per cycle. And some of our young developers said, you know what? I wish I could use Git and Jenkins and Gulp and tie them into Endeavor or these other dev ops tools or IT ops tools. They developed it as an internal use tool for a command line and we stumbled over it on accident. We said, oh my God, this is something we think customers would want. And then as we got talking with Rocket, with IBM, Rocket had a web interface. IBM had the API mediation layer and we said, holy cow, you know, this is something if we got together and we contributed, we could really start a renaissance around mainframe. And a lot of people are going, well, why? You've got proprietary tools and software. Why would you open up? Because the reality is we want our customers to find it easier to work with the mainframe. And look, I'll compete on the differentiation of my underlying product, whether it be price or function. But I want my customers to be able to tie in my software with IBMs, with Rockets, with whomever and pick because of where the value is, not because they feel locked in. And that was one of the gripes about mainframe, right? People thought they were locked in. Locked in, proprietary, weird interfaces, friction. You take the friction away. And that's not, that's your father's mainframe. That's not today's mainframe. No, that was the old, exactly. The old kind of perception. When you bring Linux and all these tools and infrastructure as code, it's just another resource on the network. Guys, thanks for the insight. Appreciate it. I love talking about mainframes. Oh my God, made my day here. So do I. Mainframes in the cloud world. Final, give a plug quickly for Broadcom. What are you guys working on? What's the big news here for you guys? Give a quick plug. Hey, I'll tell you for me, Broadcom acquiring the mainframe business is all about investment. And I mean, we're a software business, so more than 90% of my expense is people. If I'm not hiring, I'm full of it. I'm not investing. I'm hiring. We're posting like crazy. We're hiring. We're expanding to the team. And the idea is all about, there's customers have used core products for many years and they want to count on them for many years to come. We're making those investments and we're going to continue to invest in the new capabilities to help make them more efficient, effective on the platform. Your thoughts? You know, I mean, I think that, you know, that's, you know, it's interesting. If you look at Broadcom and a lot of people don't know, you know, what's their focus, right? They're not traditionally in the software space and so on. They are now. The first thing, well, they are now and one of the things that we're doing is if you look at our investment rate in R&D in general, it's up there. I mean, world class, you know, if you look at the largest, you know, most successful cloud players, forget about your large cap tech companies and so on in terms of the percentage of revenue they spend on R&D. We're far above that. We're at a very high level. We're going to continue to invest in a lot of innovation, you know, AI, machine learning, DevOps, and of course, you know, key areas of security. It's a cultural shift. You can see vinyl records making it come back. Now you got mainframes back. How much can I get a mainframe for? If I want to be the new cool kid on the block. Well, you got to go to IBM for the hardware. But I can talk to you about the software to help you with it. You got a mainframe for theCUBE, just so we have one in our house. Thanks guys. I'm going to appreciate it. I'll start with you guys. Cube Coverage here talking mainframes of IBM Think, Software Linux, The New World, Cloud Data AI. I'm John Furrier. Stu Miniman, back with more coverage after this short break.