 Live from San Francisco, California, extracting the signal from the noise, it's theCUBE, covering DockerCon 2015, brought to you by SiliconANGLE Media, with special thanks to Docker. Now your hosts, John Furrier and Jeff Frick. Okay, welcome back everyone. We are here live in San Francisco for DockerCon 2015. This is theCUBE SiliconANGLE's flagship program. We go out for the events and extract the super news. This is day two of two days of coverage, Walter Wall, DockerCon, John Furrier, with my co-host Jeff Frick, with SiliconANGLE, and our next guest is Jason McGee, IBM Fellow, VP and CTO, Cloud Foundation Services with IBM. Welcome to theCUBE. Thanks, good to be here. I'm great to see you. So CTO, Fellow, that means you're a distinguished expert. That's right. That's IBM language. When you become a Fellow, you've made it. IBM speak, yes. Look at a tenured professor. You can't mess with me now. We have good domain. We'd love the Fells on theCUBE. I want you to get your technical take. How real is Docker from a mainstream perspective? Obviously, the vibe here is solid. It's in all the conversations, whether you're at OpenStack, big data, Watson Analytics, software developers, people talk about containers. What's your take on it from an evolution standpoint, technically? Yeah, my take is we're right on the cusp of it being really real. You know, like, I think, as I talk about a little bit this morning in the keynote, I think everybody over the last year has learned what it is. Everyone intends to use it and we're right on that cusp where we're starting to deliver the capabilities needed to take it to production and to use it for real, right? So I think it's right there. You know, I think there's a little bit more to do. A lot more to do. We had only one guy who said, I'm really using it in real production. Right, right. I mean, a little more to do to get us over the line where those things can start and people can really get going on production. But I haven't talked to anybody in the last six months who isn't working on, how does this affect me? How do I factor in architecture? And all of them want to go to production, absolutely. So the conversations are elevated from tire kicking, POCs to what's the production use case is. How do I get there? How do I get there? Which is the tipping point. Right, absolutely. All right, so you guys have a huge customer base IBM. So, you know, one of the things that we're seeing in the software business is architectural shifts. And these happen once every few decades. You know, mainframe, mini, these things. What's the big architectural game changer of the cloud? That levels the playing field and or causes innovation disruption? Well, I think there's a couple things that are going on right now, Ron Claude. So one is microservices, right? And this idea that architecturally people are really starting, not so much microservices about the technology but about the culture change that's happening, right? People are really focused on like, how do I deliver at much higher velocity? And I think the combination of those ideas with cloud as a delivery platform with technologies like Docker that start to fill in the kind of technical enablers to get there is really changing how people think about building applications. I think the other piece that's really pivotal right now in driving this whole adoption is the whole notion of hybrid, right? That people really recognize that they're going to have to run stuff in the cloud, on premise, that where they run it today is and where they're going to run it next week or next year. And so they're looking for ways to kind of build that next generation of applications with enough flexibility built in to adapt. But it's not a mutually exclusive scenario. On-prem is still going to be around for multiple years. That's pretty clear. Cloud has got economics that you cannot deny. So the question is, what software runs on which and that's the big conversation like, I don't want to have the version of software from XYZ vendor. And I got another vendor over here, say it's IBM software. I'm running Amazon or something else. So this is kind of where the battleground is. For software portability, code, if you will. Explain that concept. Because a lot of people aren't getting their heads around that. Yeah, so I think you're right. I think there's a couple of dimensions to what people have to think about to make this real. I mean, first is, as an application developer, I think the center point should be, I'm building applications. How do I move those applications? And how does the infrastructure that I choose on-prem or off support that? You need a way, first and foremost, as a developer to package up the code and deliver it in a way that doesn't have to change as I move it between God environments, different hypervisors, on-prem, off, whatever. And running on some sort of hardware at the end of the day, right? Right. Whether it's Flash, memory, or Flash Drive. It's got to run somewhere. So that's where I think container fits into the picture. The second piece, I think, is no software today is built completely self-contained collection of components. In fact, one of the challenges I think the Docker community has is to open up their thinking around not everything is just a collection of containers. That you're going to have containers. You're going to have services on the cloud that you just call and use as APIs. You're going to have existing systems that you go back and connect to. So how do you make those services and APIs you're using available in a consistent way across on-premise and off-premise systems? And then the third piece is just what's the delivery process around? How do you do DevOps? How do you do delivery and things like that? You got to get to some consistent model for how you deliver and what the people processes are on delivering those things into these different environments. But you're right that we're not going to get, this is not like a temporary transition from state A to state B. I think the only thing IT's taught us is nothing ever completely goes away. Everything adds to the past and containers and microservices and all these things I think are the same. How do we have them compliment the things that we did before? Well, our Wikibon research team identified three areas that customers are interested in. You know, outside of the speeds and feeds, they have different linguistics. They say, I have new apps I'm deploying, completely clean sheet of paper. I have existing apps, I got to modernize and I want to migrate stuff to the cloud. So it seems to be the top three conversations that we're hearing. Can you talk about the dynamics on those relative to the cloud? Because cloud will be an enabler on all three. So, because whether it's on-prem, existing, the software matters, right? So I can't have cloud software running differently. So talk about that, those three dynamics. Yeah, so I think one of the ways that we've looked at that problem is to say really what you need is a platform that delivers across kind of a spectrum of delivery models that is a real trap in thinking that one approach is going to solve all of those problems. And so you need a platform for rapid development and that's where container certainly plays a role there, what we've been doing around cloud foundry and that approach plays a role there. You have the modernization space. One of the things we've been working with our customers, for example, is how do you take Java Webster-based applications and bring them into the cloud model. Sometimes that's of the shape of I want to kind of extract a set of services, do some refactoring and modernize that application. Sometimes that's the third case, which is just I want to just lift up what I had, run it in the cloud. I don't want to change a lot about it. I realize I'm not going to get all the benefits. What I am going to get is kind of cloud operational models, more elasticity, automated deployment, more efficient resource management. So I get kind of caught operations for existing things. And no one approaches the right one. It depends on what you're doing and you want a single consistent platform. Technically they should co-exist, right? I mean, I want to move a workload into the cloud. I don't want to have to do anything. I should be able to just kind of just move it out there. And that thing that I move might be sitting side-by-side with some brand new application on building and they have to talk to each other. So you want really a consistent platform. And I think that's one of the areas where the industry is still maturing and you have certain people are great off-premise, infrastructure, other people are great on-premise, some have a good pass. But you really need that full spectrum if you really want to address the broader landscape of applications that people are dealing with. Yeah, we're short on time. We've got somebody waiting in the way. All right, final question I got to ask because we love talking to IBM. We had a great event last week at Spark Summit. Great conversations, a lot of stuff going on in memory. So there's a final question. Hardware, software, we're in a software-led infrastructure now. And so software is the key, that's the value. Cloud is not a race to zero because the value's shifting. Marginal economics still apply if you do it right. So what's the deal with hardware's role in this software dev house role? Because at a surface you can extract away the hardware. Should it be optimized for performance? Oracle certainly says, hey, we will run Oracle up and down to the end to end. But not everyone has Oracle everywhere. So that kind of teases out what Amazon's shown to end to end. So is there an end to end future where hardware gets smarter? Yeah, absolutely. I mean, you know, the great mystery of the cloud, right? That all this stuff actually runs on hardware somewhere and that hardware matters. And I think there's a lot of, you know, I think maybe the balance is shifting where we're trying to build from the application down. And so you want models where the application can be simple and portable but where the optimization of the hardware can show through. I mean, a simple example, the container services we're doing on Bluemix are sitting on top of the bare metal capability that we're running in SoftLayer. So we can actually get bare metal performance out of an application running in a container on the cloud without the developer having to kind of deal with the complexities of managing a bare metal system. So there's a nice balance there. But hardware and hardware optimization is going to play a role for the, you know, forever. Certainly the flash and memory stuff is going to be. Flash and memory, processor upgrade, GPUs, all that stuff makes a huge difference. So would you agree we're at the beginning of this front wave? I mean, just early days? Yes, I think so. I think we're at the beginning of kind of thinking through how do we now introduce hardware optimization ideas back into shared multi-tenant cloud environments, right? So there's going to be a before cloud and after cloud. You know, BC, we're one BC right now. One AC, after cloud. Some people think the cloud game is won. The cloud games just started. There's so much more to do. It's totally just started. But the numbers are clear. Oracle had their first quarter of their quotes. Cloud business, no matter what you believe, the numbers are coming out. You guys are shifting 100% behind cloud. Absolutely. Yeah, so good stuff. Great stuff. Great to chat with you. Thanks for sharing. Thank you. We'll be back with more live here in San Francisco right after this short break.