 Live from San Francisco, extracting the signal from the noise, it's theCUBE, covering Oracle OpenWorld 2015 from Studio C, brought to you by Cisco. Now, your host, Stu Miniman. Welcome back to theCUBE. I'm Stu Miniman with wikibond.com and you're watching theCUBE here from Oracle OpenWorld 2015. Getting close to the end of day one. We're going to be here for three days. We've been talking a lot about systems and one of the components of systems is the Cisco UCS. Happy to have on the program for the first time. It's Girish Kulkarni, who's Senior Product Marketing Manager with Cisco. Welcome to the program. Can you tell everybody just a little bit about your role inside Cisco, what you do there? Sure, hello everybody. So I'm Girish Kulkarni. I'm part of the Cisco's data center and virtualization team and I'm part of the product marketing team and I'm responsible for the UCS marketing. That includes also the performance amplification. So essentially all the benchmark performance related amplification as well as marketing of all those results. All right, so we were talking off camera about there's big news last week. It was the Dell AMC announcement. In my career, there are a few things that when they come out are really big announcements. And I think many people missed it when UCS announced, you look back now and you look at the ripple. Companies made acquisitions, partnerships changed, ecosystems adjusted. This wasn't just Cisco saying, oh yeah, we're going to sell a server. It was transformational to the data center business model. When you look at UCS, what does it mean to you? What does it mean to your partners and your customers? Well, when it comes to UCS, essentially the kind of ecosystem partners that we have and the integrated infrastructure solutions that we have created. We are not dependent on a specific vendor or any specific partner. We are basically working with everybody and we have solutions with everybody. So essentially all the news and acquisitions that are happening around us shouldn't basically affect us seriously. That is first thing. The main component I think that is helping us to develop the tremendous traction, market traction that we have developed in the market with Cisco UCS and the integrated infrastructure solutions is essentially the performance that Cisco UCS delivers. And I would say that each and every vendor, each of our competitors has access to the same Intel processors, accessories processors. And it's only the fabric-centric architecture and the architectural innovations that we have with Cisco UCS that brings out the real power of the Intel X86 processor, which allows Cisco UCS to perform way much better than what our competitors can achieve in that. So I think in addition to that, the overall Cisco UCS architecture with the Cisco UCS manager, which provides the integrated policy-based management that allows you to bring up, configure, and deploy enterprise applications and related services in way efficiently when it comes to comparing to other vendors system. And that is where I think the difference is. And that is why over the last six years, since the introduction of Cisco UCS, we have achieved more than 100 world record benchmarks on industry standard performance benchmarks. All right, so when you look at performance and we're talking about the application space, how much do you have to tie in your storage partnerships? How do you build those solutions? And how does that go about building that entire solution from a performance standpoint? Well, when it comes to kind of configuring the systems for the benchmarks, we pick and choose the storage partners. And it basically varies from benchmark to benchmark. We may have a different storage partner while running a database benchmark versus a middleware benchmark. But I think irrespective of the storage that we use or the storage subsystem that we use, I think it goes again back to the architectural advantages and the overall design of the Cisco UCS, which drives the excellent results that we get on all kind of benchmarks. As I was saying, I mean, we have benchmark results that encompass the whole spectrum, starting from the best CPU performance, the best database performance, best middleware performance, best enterprise application performance, HPC performance. And just recently, last year, in fact, TPC Council came up with a new benchmark for big data, and Cisco UCS was the first platform and Cisco was first vendor to come up with a world record performance record result on that big data benchmark, that is called a TPCX HS benchmark. So we are always there in the forefront delivering the best results of the latest and greatest benchmarks. All right, so Gourish, when it comes to performance, of course, the challenge you have to run it, we'll look out for is bottlenecks. When I've heard people describe UCS, it's about balance and it's about understanding the applications that I'm going to be using. Could you maybe share with us how you think about that and how do you make sure that you're getting the best utilization out of your environment while tuning for the best performance? Well, when it comes to, can you just paraphrase the question again? So it's getting the best performance, it depends on what the application is, as well as what resources, I think on the networking side, there's sometimes limitations in the architecture, so how do you make sure you get the best utilization out of your environment while tuning for performance but not having to spend 10x the dollar amount? So balance and kind of cost performance and utilization of the system. Right, so that is one of the key things that helps Cisco UCS to achieve the performance that we achieve because Cisco UCS unifies your all kind of system resources, including compute, networking, storage access, and most importantly, management resources into a single cohesive system and provides the foundational automation that allows you to provision and reprovision these resources based on your application's requirements. And you can provision and reprovision these resources and tune up the system like within minutes, not within hours or within days, like with some of our competitors. And that is basically drives the performance really high because you can quickly reprovision the system, reprovision a specific asset, you need additional memory, you need additional processing, you need additional networking power, you need additional bandwidth, you can quickly reprovision all these resources just using the Cisco UCS Manager on the fabric and you are good to go. And that basically drives the performance higher compared to some of our competitors systems. All right, when the UCS was first launched, the target of the solution was really virtualized environment, specifically VMware. Can you talk about how important is virtualization and what do you see as kind of the next steps beyond virtualization or where does that fit in? So you're right, when UCS was introduced in 2009, virtualization was a mature concept and we were talking about cloud computing. So we had that advantage of designing a system that was very well optimized for virtualization because we were familiar with some of the issues that our customers were facing when it came to virtualization, as well as we also took into consideration the requirements for the cloud environment. Now with virtualization behind us, not exactly behind us, but with us, and cloud with us, the next thing is definitely the composable infrastructure. And by what we mean by composable infrastructure is what we offer with the Cisco M-Series modular service where essentially you can provide the capability to use the infrastructure as a code. By offering the capability of infrastructure as a code, we mean providing the capability to customers to create servers, or you can call them composables servers or composable infrastructure with the existing pool of resources that you have. That could be bunch of processors, memory, local disk, local storage, and SAN and IO subsystem. And essentially these kind of composable machines is the future now because with virtualization you can only do so much. You have one virtual machine and you are slicing and dicing that same machine, that same pool of resources into multiple things and getting things done. It worked for some of the traditional applications, it worked for some of the existing applications, but some of the new applications that are coming in, like big data analytics, like IoT, internet of things, internet of everything, for those new applications which are being developed within the data center at the edge of the data center at the outside the data center everywhere, right? For those kind of applications, you need underlying infrastructure that essentially can be easily deployed, quickly and easily deployed. It is operationally simplified and most importantly it is economically efficient and that is what the composable infrastructure is going to be and that is where we are going with the M-Series modular service. All right, and if I hear you correctly, that's something we could do programmatically, correct? Yes. Okay, one of the biggest challenges we always saw is somebody go to the network team and say, oh hey, I need you to adjust this. Can you fix the quality of service? Can you do those knobs? And it was like, no, we're not going to do that for you. So we need to be able to give tools so that there's automation, there's ways I can manage things. Definitely, right from the beginning when UCS was introduced, it was a policy-driven architecture. We were way ahead of our competition and we just basically continued evolving it and revolutionizing it to reach where we are today. All right. Excellent. Thank you so much for coming, sharing your viewpoint with our audience. We'll be right back to wrap up day one here at Oracle Open World on the exhibition floor. Thanks for watching. Thank you very much.