 Live from the Sands Convention Center, Las Vegas, Nevada. Extracting the signal from the noise. It's theCUBE, covering HP Discover 2015. Brought to you by HP. And now your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are here live at HP Discover 2015. This is theCUBE, our flagship program. We go out to the event and extract the signal from the noise. We're here at HP Discover for the sixth year. I'm John Furrier, my coach. Dave Vellante, our next guest is Paul Durzen, VP of Product Marketing, Converged Data Center. Welcome to theCUBE. Great, thank you to be here. So you guys had a big announcement coming on. Converged Data Center's hot trend. Software-defined data center, software-led infrastructure, software-defined infrastructure, software-defined storage. Software is really the key to success these days. However, you still need hardware to run stuff on. Absolutely. Whether you're in the data center, in the cloud. Tell us what's going on. What's the announcement? What is it that you guys are releasing? Absolutely, thanks. So we're here today to talk about project synergy and composable infrastructure. So essentially, project synergy is our roadmap or our project we're putting around to deliver composable infrastructure. And as you said, everything really is about software. The future's moving towards software. And when you move to that world, what you want your infrastructure to be is to be composable or to be software-defined. We'll call it infrastructure as code. So you want to be able to go and drop a personality onto that infrastructure and have it become whatever you want it to be right before runtime. So project synergy is a first look now. So you're essentially releasing this. Yes. Composable infrastructure. Sounds like, I love infrastructure as code. It's the term we use in the DevOps world. Yes. What is it? Why, I mean, go deeper on what, I mean, so that's the announcement. What actually is it? Is it software? Is it hardware? Is it a solution? Is it a mix and match? What is it? Great, great question. And I love your tie-in to DevOps because when you hear about what we're talking about here at Discover, we talk about the new style of business, right? And that's really the DevOps world. That's the world where IT is transitioning to making money and really becoming a value creator. And yet most of us here are in this traditional world where we live in a world where making a change is really hard, where we have big teams, process-oriented environments to make a small change and where it becomes really, really difficult to do what we have to do very quickly. So we essentially keep our environment as static as possible. And these are two different worlds in the data center. The DevOps world where things are moving, continuous change in the traditional environment where actually change you don't always want to have happen. So what we're doing with composable infrastructures, we're bridging the gap between the two and we're creating a common infrastructure for the traditional so you make it a lot more efficient and for the new style of business so you can actually move to the new style of business. And what that allows the traditional enterprise to do is become more efficient and again have investment protection so they can move to that new style of business. You know, five years ago I asked Pat Gelsinger who came from Intel to be an executive EMC on theCUBE and I said, you know, being an Intel guy I said, hey Pat, this is more of kind of a personal question. I said, what comes first? Infrastructure or the app? Of course what he said, oh, infrastructure without the infrastructure the apps don't exist. But what you're getting at here essentially the apps are the centric of all the focus. Absolutely. And you want to have an infrastructure that can be foundational but yet adaptive to whatever the needs are for the app, right? So the business you might have a workflow app here, an HR app or financial vertical where the apps need to be dynamically provisioned or dealing with resources. That's exactly right. So when you think about a composable infrastructure it can be optimized for the application right before runtime. So that's exactly right. You know, whether you're doing a cloud native app or a traditional app you have the ability to very quickly instantiate that service and very quickly optimize the footprint for that application. So the hot buzzwords right now you mentioned cloud native apps, it's hot. But a lot of people haven't really built a cloud native apps that's still coming. That's correct. Another buzzword that's hot out there is microservices. Microservices, that's right. Yeah, I love that. What the hell is that? Okay, what this teases out is this new style of development if you will, right? So containers have been all the rage with Docker and what not and certainly OpenStack this year we were just at OpenStack so I'm interviewing a lot of the HP folks on the cloud side but OpenStack is getting a lot of steam right now and the commentary at OpenStack was containers aren't really moving that fast and the thesis wasn't the observation from like Lou Tucker and a bunch of other guys was there's no native apps out there so it's coming. So there's some, I mean not that it's build that they will come they're definitely coming so tease that out it's evolutionary. What do the customers want? Talk about that dynamic of it's enabling but the apps aren't yet there. Exactly, so just like you said cloud native is a very small percentage but people know they have to move there and right now they're in a position of building a cloud native platform or a traditional platform two different investments two different worlds and there's no possibility to move in between the two so we figured if we can build this composable infrastructure bridge that gap what we do is essentially we create these fluid pools of resources so you can take these pools of resources optimize it for the application we put on software intelligence so you can define that again that infrastructure from software right before runtime and then we have this unified API and the important thing about what you just said is this unified API allows you to start to take this workflow and align it to the infrastructure in a really easy manner because the unified API will cover the full infrastructure and is your point of automation. So what about sort of SLAs through that API? I mean can you talk more about sort of service levels that I can program thresholds that I can develop and how programmable is that infrastructure? Well the goal of that API and that infrastructure is to hook into whatever higher level orchestrator you're using so for example obviously we believe that HP Helion is a great way to go it provides a nice vertical stack from the developer platform through you know open stack cloud management into the infrastructure in this case it would be into composable infrastructure so you're able to go and take whatever higher level orchestration tools you want and start to program this infrastructure and all of that happens through your existing set of tools. Okay so when I think about this discussion I think about a customer wants to build a cloud internally and potentially connect to some kind of external cloud that's sort of the play here is it not? So that's absolutely a possibility what our goal is really to make the creation of the underlying infrastructure for that cloud as easy as possible and hook into whatever tools are doing that orchestration or that procuring right so the goal is when you look at these higher level orchestration tools there's a couple things about them number one normally you need a lot of them to create your automated environment and number two normally you need a lot of very educated people to go and build those services and to program that infrastructure what we want to do is we want to allow the infrastructure person to retain control of the infrastructure by creating these personalities of the infrastructure serve them up to the higher level orchestrators so when the orchestrators go to put down that application that application knows exactly what to call the infrastructure person stays in control of the infrastructure and the DevOps person stays in control of the app because you don't want one doing the other the other doing the other because that way you get a mess So it's not a one size fits all you've got this personality notion can you give us some examples of the types of personalities that you're seeing in the marketplace where there's demand Yeah absolutely so the goal here again is composable infrastructures about creating these fluid pools of resources defining these pools of resources through intelligence and having the unified API so for example you can imagine being in this world where you have an ERP application and that ERP application may be very compute intensive so you need a compute intensive footprint for that application so you can create that from the infrastructure on the same time you may have a mobile application or some type of say a different type of application Hadoop application that needs a lot of disk again from the same infrastructure you can give a compute intensive application compute intensive infrastructure a disk intensive application disk intensive infrastructure and I'll do that in an automated manner And what am I as a buyer what am I buying am I buying a box with some software am I configuring that I'm interested in what the channel wants you to do because the channel might want to use something different somebody's server somebody's storage what am I buying Yeah absolutely so our composable infrastructure is essentially a set of infrastructure hardware with software on it and the software provides that intelligence and the API and the infrastructure is obviously the foundation of it all and again our goal is to give you the best end to end experience so you could take that infrastructure block and view it as being a composable block hook it into say HP Healy and OpenStack HP Healy and Dev platform have a complete end to end solution but again we know customers like choice so you can hook into VMware into Microsoft into Docker into whatever higher level constructs you need to hook into So is it a single skew or is it an open sort of choice is it something in between Yeah so you'll see choice you know we have a lot of different options on how we skew it up so to speak but really the way we view it is view it as a set of composable infrastructure with a layer of software intelligence on it that you can drop on to that infrastructure to make that infrastructure So I'm basically going to menu and I'm picking from that menu that's what I'm essentially developing a solution for my needs You're creating a set of infrastructure personalities for the application so when the application has to go on to the infrastructure it takes that personality plops it on creates it and then goes there So the channel loves this Yeah because it gives them choice Absolutely But others might say well but I just want to keep it simple Well yeah that's a great point and actually that's the whole goal of composable infrastructure because when you look at the environment today for example if I go and I want to change a piece of firmware on the NIC I'm going to bring in my server person I'm going to bring in my network person I'm going to bring in my storage person my virtualization person my application person just to change one little thing So it's a very difficult in a very painful environment and the interesting thing is when we solve for the DevOps world which is continuous change you actually enable those people to work very very efficiently So what it does is it creates a very easy and efficient environment in the traditional and because we've already created that for the DevOps world so to speak Paul I got to ask you about the DevOps which is one of our favorite topics Yes People know we love the DevOps It really is one of the most amazing innovation areas right now in technology business All the actions sending around DevOps which kind of touches you know essentially the pillars that you guys are rolling out Yep It's wonderful right It's great It's enabling more compute which employs more big data in memory A lot of great action there But what's interesting is you're talking about composable infrastructure Yep You got now two dimensions You have horizontally scalable which is an open source kind of concept on commodity hardware which you could argue industry standard hardware is basically HP commodity industry standard you know boxes versus vertically integrated purpose built solutions Yep So this is not a mutually exclusive situation Can you explain Well that's on my belief Yeah But do you agree that they're not mutually exclusive What is the scenario going forward for a customer who has to put their eye on the 20 mile stair on the future and say Yep I got to build out the next generation architecture Do I have to pick one Can they both work together because I think I hear you saying is it doesn't matter You want to go horizontally scalable that's cloud Exactly You want to go vertically scalable That's That's exactly right So the interesting thing today is you have a bunch of optimized let's call them servers in the case of infrastructure for applications The goal of composable is really to create that optimization in software So now what you don't need is you don't need these very specific products Instead you create these pools of resources and you just pull from these pools of resources So that underlying infrastructure you want it to be reliable Right We keep on hearing that from customers They contemplate going to white boxes and a lot of enterprise customers who have gone to complete generic white boxes have come back because they have reliability issues and they're not happy So Management nightmare too And it's a management nightmare It's a certain scale point It looks good on paper You next thing you know you're out here you go Well the scale points change Right I mean it used to be proof of concept now you're up and running in production and then it's breaking down Exactly I mean that's pretty much the cloud business right Yep Spin stuff up easily Lower cost You see some value but then the double down part is scale Exactly So we want to do both Like you said We want to make the management easy We want to bring in the reliability But we also again want to create these fluid pools of resources that we can define and be optimized for the application and that way you get exactly what the application needs because another thing that happens in the data centers we have stranded resources right we're forced to buy a footprint with what I'll call fix ratios certain amount of disk certain amount of memory certain amount of CPU and if you start to run out of one resource you have to buy a whole set of resources again to get what you need Our goal here is to have it disaggregated a key part of creating these fluid pools It's a disaggregated architecture so you can get the exact application footprint So you're the VP of product marketing I mean you're in a hot area you excited about your job I mean you must be pretty pumped Oh I love it This is a future I think You guys probably remember when the Converged Data Center came around To me this is the next step of the Converged Data Center I think we have a great opportunity to take a leadership here What's next So you're announcing synergy First phase What's your go to market What are you guys doing after the event What's happening next Is there outreach What's the plan Yeah so the plan is like you said we have phase one here today which is around that composable API or the unified API we have a partner program around it we're integrating and then going forward you're going to continue to hear exciting news we have a good roadmap and that's going to continue to roll out I'm sure next time we talk we'll be So it's a product development kind of milestone so it's not so much you go to market bring to partners yet It's a product development It's a product development It's absolutely So one of the philosophies that we have is we want to make sure that as you move to the future we preserve your investment as much as possible So these open APIs you can use them today you can put them on today's infrastructure but they'll translate as you move to the future and they're essentially the roadmap to the future So you guys will engage with customers on this product Oh absolutely and sell it And whose services your organization or which group Yep our organization can service it partners can obviously create their own around it you're going to do marketing in market Yep And are you seeing sort of expecting to see early uptake with cloud service providers small size companies large companies help us break that down Well so this is really to me this is really an enterprise play So as I said I think we've solved some really significant problems for the enterprise and that is in the traditional environment it's really hard to move and they know they have to go to the future and build two infrastructures and they don't want to do that So where we can do today with the traditional enterprise make their enterprise more efficient and when they have to move to the future they already have the infrastructure and the setup to move to the future So I really view this as an enterprise play I think it's going to be a great area in the enterprise Alright Paul well thanks for launching on the Cube we really appreciate congratulations we love DevOps it's a hot area of course we'll be following you guys and congratulations it's super exciting horizontally scalable vertically scalable it's all happening software defined infrastructure it's the Cube we'll be right back bringing you more signal from the noise go to hpdiscover.social our new engagement site that puts all the live feeds there all the crowd chats go to crowd chat we've got a big influencer group right now streaming out of chat and again stay with us we'll be right back after this short break with our next guest