 joined by Satya Nadella. Tell us a little bit about your thoughts on this event in general. Last year was about big data. This year it's a little bit more focused, a little bit broader focus on the modern enterprises. They say, what's your take on kind of this event? It's a great event. This is my first time here as well. And having a chance to even see a couple of panels and just participate. I think this notion of a modern enterprise is for real. I think that there is a re-imagination of what does infrastructure mean? What do applications mean inside of the enterprise? And we're going through this tectonic shift, which we participate in. And so to have a forum like this to discuss that, it's just great. So let's dig into that a little bit. What makes the modern enterprise? It's certainly cloud and virtualization. You've got the big data piece, kind of the DevOps model of application development. How do you kind of define what all bring together all these different elements? What makes a modern enterprise a point of view? Yeah, one of the things that I like to sort of make sure we focus on, I work on the infrastructure business at Microsoft. So if you're in the infrastructure business, the key thing is to be in touch with the applications. And it turns out in our own case today, we're building a pretty diverse set of applications, both consumer and enterprise. So we're building Bain, which is an applied machine learning application. We're building Office 365, which is an enterprise-focused collaboration communication application. We're building Dynamics and another enterprise, CRM, ERP in the cloud application, and what have you. So that diversity of applications makes you rethink what is the infrastructure needed from storage, compute, as well as the network. And so we're building a new operating system for the modern enterprise to be able to deploy these modern applications. So that's kind of how I conceptualize it. I would say there are four major elements to it. The first one is inside of the data center, you have much more of a software-driven data center where you're orchestrating your compute storage and network in support of your applications, either at the data center or multi-data center scale, because there's not a single enterprise that's not using some public cloud provider or another service provider, in addition to what they already are virtualizing inside their own private cloud. So that is all a software control plane. And so we're really thinking about what is the modern operating system that enables you to manage the data center. A second dimension would be what is driven through consumerization of IT. I like to describe it as transforming IT to be much more people-centric. So you want end users to adopt the devices they want and still have access to all their applications and data. And yet, IT needs to be able to set compliance and policy, and so how do you really reimagine that is another dimension. Big data is something you reference. There's not going to be a single application that's not a big data application. And so those are the major, major themes. And the last thing I would say is this DevOps. So not only have you built the application, but it's even the life cycle around the applications being reimagined, how developers and operations professionals come together in support of an ongoing improvement and continuous integration. These four mega trends, I think, constitute a modern enterprise infrastructure. Interesting. So let's dig into a little bit about what you mentioned about the use of kind of public cloud infrastructure, as well as your internal data center. So you've got these hybrid environments that are starting to emerge. Again, pretty much software-led infrastructure is what we're calling it, a Wikibon. How do you go about actually making it possible for CIOs and their teams and to actually manage those environments as efficiently as possible? Making decisions about which applications are deployed in the public cloud, which are deployed in your data center, how they interact, potentially applications that are drawing on data from both spots. It's obviously can get very complex. So Microsoft is one of those public cloud providers with Windows Azure. So how do you approach that problem? So if you sort of take what you just described, which is if you sort of start with the design point that there will be a public cloud, there will be a private cloud and a service provider cloud, then how you think about the software control is going to be defined by that design point. So it's not going to be narrowly defined as bring everything into my data center and I'll help you manage it. But it is actually distributed. So I think of this as the true fruition of distributed computing and we believe in that. So then what are the things that matter? First is identity. So anything, whenever things get distributed, the most important thing that brings back things together is actually identity of users and identity for resources. So Active Directory was a great resource for many enterprises in terms of how they tame the complexity of the previous generation of client servers. Now we've replumbed and reimagined Active Directory with Azure Active Directory. So this consistency in directories helps IT administrators manage this complexity. The next one is virtualization. So not only would you be able to virtualize on your private cloud, you should be able to move the same workload which is virtualized to any of these other clouds. So you need this degree of guarantee that the performance characteristics of a virtualized workload get maintained across all of it. So that's another thing that with our Hyper-V investments and our Azure investments we are making sure that happens. The other one would be management. So if you can be sitting on the system center management console and the orchestrator and looking at a workload which could be in fact in one of these clouds or in fact the tiers of a single application could be split which is the front end is on Azure, the back end is back in on-premise. And so that's also very, very important. They have a management tier which is the control plane that allows you to manage this complexity. And lastly it's the consistency of the application platform itself. So if you're building a development you never want to be in the state where you build a great app but you can never check out. So if you build it in the public cloud in the case of Azure you should be able to take it and run it on a private cloud or on a service-provided cloud. So these four things are on identity, management, virtualization and application platform. I think it's the core investment you've got to make to help enterprises truly adopt the cloud while it's complex but you've got to tame the complexity. Then of course what you're talking about really is a lot of data being generated. Companies of course want to start taking advantage of that data, they want to analyze it, they want to actually take those insights and turn them into either applications or perhaps convey them to executives and others in terms of visualization. And of course one of those underlying platforms is Hadoop. Talk about Microsoft's approach to Hadoop. I know you're working with Fortinworks. You actually kind of discontinued working on your own big data technology. When you realize I think that Hadoop is going to become the de facto standard. So talk about how you're making it possible to bring Hadoop into this environment where more and more companies are looking to bring that in. Maybe as a big data hub, kind of store a lot of data and then feed it out to different applications, different workloads. What is your approach to actually making that I guess enterprise ready? And making it easy to get started and then turn maybe science projects into really production level of acquaintance. I mean this notion of being able to take data and convert it into insights in support of enterprise goals is sort of the holy grail of this moment. And so one of the things that we are actively doing is to bring a lot of the traditional value we've always had. If you think about the momentum we have with our self-service BI capabilities on the edge of data, which is Excel, SharePoint, SQL analysis services, is where all data goes to in order to be able to drive insights with end users. Because at the end of the day humans will be involved to be able to drive insight out of all of this data. So now the question is how do we take that edge loop and connect it with the information production which is upstream. And that is where we're completing the story with having HD insight, having even a relational interface on top of HD insight for in-memory ad hoc query analysis like a data warehouse on top of it, which I think the Hadoop community itself is adopting, which is a SQL interface on Hadoop is probably one of the more talked about things nowadays. And so this notion of having a complete data platform, everything from MapReduced to stream processing to SQL like query interactively, and then empowering end users and workflows with data around end users with SharePoint, Excel, where we've invested in things like power pivot and power view, which are actually powerful in-memory databases. In fact, I would say the most powerful in-memory database now is power view inside of Excel from where you can issue a SQL, I mean basically a Hive query to HD insight and populate millions of rows in a tabular column form that you're very familiar with. We think that that democratization of big data is going to be very, very important to acceptance of it as you said it from science projects or just being in the data science department to being ubiquitous. So we've only got time for one more question. So we'd just like to get your kind of future outlook. What are some of the key priorities for you and your group over the next, say, six to 12 months? I mean, the key thing for us is really bootstrapping our cloud business. We've got some fantastic traction with Office 365. It's really doing very well in the Q3 earnings. We talked about how we have on a run rate basis a billion dollars in revenue going to Office 365. I mean, many customers who are coming to Office 365 never bought an exchange server from us. So we're even, you know, it's not even zero sum really in the short run, at least. And so we are very glad with that. And there is, Azure is just a natural complement to any customer who's already got Office 365, SharePoint extensions, the end user BI, Active Directory administration. All of these are sort of very natural extensions. But Azure itself now has got very, very significant momentum yesterday. We talked about how Azure and Azure services with all of our service provider partners has also got a billion dollars of revenue for us. So that means when it comes to the core of the enterprise and their move to the cloud, which is going to be complimenting a lot of what they're already doing in on-premise, is something that we're a pretty major player on. And if anything, we want to be solving the here and now practical problems with a forward-looking vision around identity, around consistency of the management plane, around virtualization compatibility, around the application platforms. And I think that that's what we're really up to in the immediate future. All right, yeah, I think you really hit on something there with, these are going to be hybrid deployments. They're going to be, you know, just much like in big data, you know, Hadoop isn't going to come in and replace your database, your relational database. And neither is the cloud going to replace your internal data center. They've got to work together. It sounds like you guys are working hard to kind of make that as seamless of a proposal as possible for your clients. So, Satya Nandela from Microsoft, we appreciate you coming on theCUBE. Thanks very much. We'll hope you come back and join us another time. Great, thank you so much. We'll be right back from the Excel, Stanford, and Cosium with our next guest right after this.