 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, Dave Vellante and Jeff Frick. Welcome back to theCUBE, everybody. This is HP Discover 2015. Dave Vellante with Jeff Frick. John Furrier unfortunately can't be here and I'm sure he's sad. Sargillai is here as Senior Vice President of the Communication Solutions Business Unit at HP. Sargillai, thanks for coming on. Despite John up here. Yeah, I mean, it's unacceptable, really. That's all I'm talking about. You can just look right at him and tell him. So I know, so the first question, I know John would want to ask, so you're one of the architects of HP's cloud solutions. You really did a lot of work there, great work. And then when you moved over, everybody said, uh-oh, what's happening? Sargillai, and then all of a sudden, we're seeing this CSB explode. Huge group, thousands of people. This is a real deal. Yeah, well, you know, the identification of TELCO is a big opportunity for HP. And we decided to sort of put all the assets together that HP has, whether it's our CMS group, right, the communications media and services group, the stuff that we have from NFV and also now we just acquired Context Streams. So we were putting a bunch of assets together to go after this business. It's a big business for HP, big opportunity. So recent acquisition, maybe you start to talk about that, what was the play there? So, you know, we're obviously pushing very hard into NFV. One of the key elements of NFV is something called service chaining. So if you, when you're running communications, when you're running a service, it's actually usually a chain of different functions that you're doing in the service. For example, if you're browsing the internet, maybe you have video compression, maybe you have a firewall, some subscribers may have parental control, and so on. And so as you move into network function virtualization, the ability to chain these services is very important. And Context Stream has really good technology to do that. That's been proven with very large providers. And also they're doing it on top of the ODL. This is an open source solution for SDN. And so we felt that that would be a really good addition to our portfolio. So talk about what's going on in Telco. A lot of heterogeneity, they need to scale, they need to compete with the Amazon Netflix, you know, scenario that's going on. Talk about that problem and how you help. Yeah, so I mean, look, the interesting thing in Telco world is that the carriers have built very good solutions, good networks over the last two decades. And they've been very successful, right? You get good connectivity, LTE, good coverage, you get stronger internet to your house. So I mean, they've been very successful. However, increasingly they have two big challenges. The first challenge they have is that the monetization of those networks is increasingly being done by outsiders, the over-the-top players. So, you know, if they used to make a lot of money, for example, on text messages, remember, that used to be very expensive, billions of dollars a year, that's gone, right? If you are using WhatsApp, they're using Apple, it's gone. So more and more, the value add services, where the monetization is happening, is outside, people are using their network to monetize, but they're not getting the money. Now over-the-top players are getting the money. That's one problem. The other problem is that there's an explosion of bandwidth, obviously usage, we all know this, the video, Netflix, or on the phone, and increasingly they have to invest more and more and more to make the same money as they made before, even less. And so that's not a good strategy. Like, you know, if your OpEx is going up, your margin's going down, that's not a good one. Ask the publishers out there how that works. That's not, now they know this, okay? These guys are very smart, right? They're very savvy, and they've said, you know, a few years ago they said, okay, well, what do we do? We got to do something, right? And really, you know, they concluded that they have to get the technologies of the folks they're competing with over-the-top players. They have to move into sort of the cloud IT realm and leverage those technologies so they could be more agile and improve their cost structure. And that's how NFV, Network Functions Virtualization, came together. It's just a name, but it actually means many things. Which came from them. The providers actually came up with this. It's not suppliers, the suppliers are meeting it. And you know, that's about taking a lot of the functions that they have and moving them to a virtualized environment. But it's more than just virtualizing. I can talk later about the stages, but it's more than just virtualizing. It's really about adopting all the cool technology that IT and cloud has taken off over the last five, six years into the telcospace. Because remember, what they have today has been 20 years, right? And so their stuff is, you know, what was done 20 years ago. If you look at IT, you're not running stuff that we're doing 20 years ago. So in order to get that agility, they need to adopt these technologies now. It's not easy because these are big networks, complicated systems, but that's what network functions privatization is all about. So when I think of telcos, it's a lot of diversity, right? I mean, it's a lot of diversity in the enterprise. Telcos is probably even worse in terms of operating systems and, you know, software and the like. How does NFV solve that problem? So again, NFV is more of a paradigm in terms of today, most of their functions. And again, when you think about functions, it's important to understand when you're, when you, if you pick up the phone and you do a, you know, internet or something, browse the internet or look at a video, you know, routing and switching, what most people think is the key functions is a very small piece of what's going on there. There's video compression, there's IP resolution, there's caching, there's firewalls, there's identity management. So there's all these functions that have to happen in a chain to make all this come together. And historically, they were all on monolithic hardware devices because that was the state of technology when they had to put all these things up in the network in the last decade. So the first issue of NFV is, let's get, let's decouple the hardware from the software. So let's take this, you know, these are all software functions, let's decouple them. And at least, you know, we will have, we won't be tied to the slow hardware innovation and we can actually run this as a software paradigm. That's stage one. And that's, you know, that gives you some benefits. The benefits it gives you is that, you know, you now at least aren't tied to the hardware. But it still doesn't give you the ability to move very fast because you're still using the same old monolithic software. It's just not tied to the hardware anymore. And stage two, what you call virtualization is where you say, hey, you know, if I'm running one thing on a server, why don't I run five things in a server? I'll need less servers, right? Now, there's some issues there because the performance, these are real-time systems. So, you know, you have to make sure you get the performance but that improves things as well because now you get better cost, right? Less capex and also less space and power. You can buy bigger servers to solve that problem, right? Well, it's not only that, but yes, you get some improvements but still a lot of their biggest, actually, impediment is actually operational costs. And that doesn't really improve that. Next stage is really what we call cloudification. And now you're not only, you know, virtualizing but you're now running in a cloud model which means that, you know, you have automation, you have elasticity. And so now you've improved also to the mode where you have a much better sort of management paradigm. Now, there's a little subtlety, and we talked about this in the past in the enterprise, that if you're using a cloud and the enterprise wasn't designed for the cloud, you get some benefits, but if the app wasn't designed for the cloud, then, you know, it's still a static app. I mean, you can do some things with it but it's not the scale-up model. And so that leads us to phase four, which is what we call decomposition where you really look at all the different functions across your different apps and figure out how to divide them into microservices that you can pass up into services. When they get to that phase, cloudification and decomposition, that's where you can really move sort of of a DevOps model. And now, not only can they, you know, operate in a much more efficient fashion, they can roll out new services much faster. Like, if you think of your cable box, for example, like, when's the last time you got a new feature on your cable box, right? So, you know, but we live in a society where you expect to get new features every day. So when they get to that phase, then they can now really compete in terms of agility. Now, it's a long process. This is, you know, these phases are going to take some time and they're not, you know, perfectly one after the other, right? There's some overlap, there's a long tail, but this is the journey they're on. So the first couple of phases deal with my sort of, what you said before is a monolithic server problem. So take that away. It's really a hardware. The coupling of the hardware. Okay, so that's, like I said, that's fine. But it sounds like from what you described, they have a software management problem. It's those... Well, they have an operate, I mean, they have multiple challenges. And again, it's not really a challenge, it's more a question of where they have to go. I mean, everything they've built has worked really well, but, you know, like anything, at some point, there's new paradigms and new ways of doing things. And if your competitors are doing it and you're not, then, you know, they have advantages. So, I mean, the OSS-BSS, which is really how they manage their whole network and do assurances, you know, that's something that was designed 10, 15 years ago. And so it's just not agile. And so as part of moving to the cloud, they also have to change their whole orchestration framework. And that's difficult. I mean, we have products, something, a product called NFP Director that comes out of CMS, for example, that allows you to do some of that bridging. But that's part of the whole thing, right? You got to move to a software model, you got to move to automation, you got to change the whole paradigm. It's not that different than what's happening in the enterprise. But there's a lot more constraints, because for example, when you virtualize, when you cloudify, you know, the apps, the performance matters. That's why we have something called carrier grade helium where we sort of made carrier versions of OpenStack that enable you to get some of the things you still need in the carrier. You need resiliency, you need fast failover, you need much better performance in terms of what you get from your hypervisor. And so there's a lot of things, it's a workload, but it's a very specialized workload. Presume, ultimately, they want homogeneity so they can scale their levels, right? I mean, that's the Amazon vision, right? Well, it's a bit, again. I mean, again, in terms of infrastructure homogeneity, they want to get to the point where, again, if you look at a very sort of gross level today, they have a bunch of different silos. There's a silo for your mobile, there's a silo for your TV, there's a silo for your internet cable, it's just a silo of things. What they want to move to is basically have a platform and that's what we're building. And on that platform, you run apps. And all the previous things that were appliance or special hardware, just an app. And that gives you that homogeneity, the ability to move apps around, to scale up, to scale down. Maybe you want certain apps to run a certain hour. For example, if someone's watching videos or to show or something, then you run a bunch of video compression apps and you pay for them when you use them. And when you stop using them, you don't pay for them. So it gives them a whole new level of flexibility on how they operate. In the enterprise business, you have to be careful as a supplier not to over-rotate to all the new stuff. You have to open stack, you have to open stack is great, but making a ton of money off of open stack today. So you can't just go all in. I mean, I'll rotate over to open stack. You have to service the existing stuff. It sounds like the telcos want to over-rotate, but they just can't. They can't get there. I mean, they want to get there as fast as possible, obviously, but it's very difficult because a lot of the issue, like the enterprise, but even more so is organizational. Because if you want to operate in sort of a horizontal fashion, your organization needs to be horizontal. You can't have these silos, like the network silo, the server silo, the operation silo, right? The dev silo, I mean, if you want to run dev ops, the dev and ops need to be together. So now they know this. Because again, it's not that they don't know this, but you have an existing system that you have to run. And so it takes time to do that. They definitely want to move there as fast as possible. It's interesting what you said about the suppliers. Not all the suppliers want to move there as fast as possible because it cannibalizes their existing business. We don't have that problem at HP because we're not in the existing system that's getting cannibalized. So for us, we're happy to go as fast as possible with them, but it's a journey. How would you organize to get them there? And is it aligned with the way they're organized? I mean, you don't want to necessarily be aligned the way they're organized in these silos. Well, I mean, this is why the carriers buy solutions. They don't buy boxes. And this is why specifically my group is called the communications solutions group as opposed to something else because we sell a solution. We take a bunch of hardware from HP, a bunch of other solutions from HP, software from HP. We take stuff from our partners. A lot of our play is this big ecosystem that we've built because one of the things that the carriers want also is choice. They want to be locked in and we package it all up as a solution that they're going to put in the carrier. And some of the technologies we build ourselves, obviously. So it's really about solutions from their perspective. And so we've organized ourselves to be able to engage with them at a solution level. And part of what you have, you've got services in your organization. That's one of the pillars that you mentioned. Where did that come from? Was that homegrown within your organization? Did you pull it out of different parts of HP? So again, I mean, if you think about the, who they want to partner with on this journey, and this is what makes it so unique for HP, is that they want a supplier that obviously understands IT, has all the sort of generalized hardware that they would now move to, which is more generalized servers as opposed to specialized stuff, generalized storage, networking. And someone who has sort of open source cloud, which we obviously have as open stack, and someone that was in history of telco. And if you look at telco, we have a big telco asset because we have what's called the CMS portfolio, right? So HP has actually had for a long time something called CMS, which is our communications and media solutions group. And that group provides solutions for telcos. We have a lot of HLRs for registries, when you roam around the world, a lot of the time it's through our HLR, something that actually runs at six nines, if you can believe that, if that goes down, they're off the network. We have a lot of OSS and VSS management stuff that we've built, and we've leveraged that ability. So we actually have a very big portfolio, and my group has now sort of, that group moved into that group. So that's something that HP's had for a long time, but they weren't organized as part of the system. Which is very, and you've pulled that together. Yeah, we had a bunch of different groups, and as part of sort of building solutions for the telcos, we sort of put that together in one entity to go after this opportunity, because we see this, all the way, you know, Martin and Meg, and Antonio, they see this as a tremendous opportunity for HP. And you're a substantial organization, we're talking thousands of people, is that right? We have thousands, yeah, many thousands, many thousands, because again, there's a lot of service to delivery. So again, when you work with a telco, right, you have to deliver the service to them. They're not going to buy a bunch of boxes and string them together. And so a lot of my organization are actually doing service delivery, they're actually taking the solutions and applying them onsite, so they're spread over the world with the customer. What about partnerships? What are you doing there? Who are your big partners that we should be focused on? So again, it crosses many, many different elements, but if you, you know, we launched the OpenMV program about 18 months ago, and that was all about going after this NFV opportunity, and part of that is a whole partnership ecosystem, and there's different levels of partnership, because again, you have to think about all these monolithic silos now, the carriers want to consume apps. Well, who's going to write the apps, right? So you need partners to write the apps, right? We're providing the platform, right, ecosystem. So we have a lot of partners. First of all, we partner with the existing network group and providers, right? ALU is a very big partners, Val Suckertel, Nokia, even Ericsson. We have a strong partnership with Wind River. It's working with us to put together some of the carrier-grade stuff because they have a lot of history there. We have partnership with Intel. We have partnership with Rokade. A lot of the players who are going after this business, and then we have about 150, I think a bit more, ISVs that are actually writing apps for this area. And again, it's a homogeneous environment. It's a heterogeneous environment. For example, we bought Context Stream. They provide a certain solution on SDN. ALU has a solution that we sell, right? DCN from SDN, and that's very useful for certain applications. We support auto-controllers. So this is not a sort of one-size-fits-all. This is about building solutions with our partners for the carriers who ultimately have the final choice in how they want to build it. We've seen some consolidation there as Alcatel's getting acquired by Nokia, right? Is that, all right? So. Yeah, they're both very good partners, so all good. So that's fine, they're good there. Now, when you talked about microservices before, who, what role do you have in that? Is it the platform to build those microservices on top of? Do you actually? Yeah, so we like what happened in the cloud. Again, this is not something that we came inventive of. It's really, you're looking at the cloud paradigm. And one of the reasons sort of I'm driving this is because I have 20 years of networking and many years of cloud. And so this is really a cloud and networking come together. And so it helps sort of living on both sides of the house. We believe that over time, right, there'll be more and more services as part of the platform. And so we will be providing some of them, but some of them will come from partners as well, right? You can see a library of service coming from partners. But we believe that increasingly, a lot of the apps might be a bit thinner and the services will be thicker, right? If you look at some of the cloud apps that we were building, even mobile apps, right? The apps are no longer thick as they used to be and that gives you the agility. So we will be building some of that capability, but we will also be leveraging partners for that as well. And what about developers? Developer outreach, you know, we furrier. It's always talking to whoever, whether it's young John's or other parts of HP, critical lever, what are you doing there? Yeah, so we're doing a lot there actually. Again, part of the thing that the carers want to do where we're very active is if you look at what's happened in that space, let's say on the last few decades, the barrier to entry was used, right? And therefore you saw very few suppliers in Telcos because you needed $100 million before you could actually show up to Telco, run a park, do something, and no VC is going to fund that or when you can go do some web thing. And so, let's say it'd be successful, at least you can believe it, right? And so one of the whole thing, purpose of NFV is to create an environment that fosters the ability of much more innovation and much more competition. And again, part of the open NFV sort of program is labs around the world that enable our partners, and again, partners, ICVs, they don't know, it's not exclusive to come up and without investing anything in the hardware and run their apps on our platform with the carriers we have around the world, we have in Europe, we have in Israel, we have in Korea, now obviously we have in the US as well, and to be able to show their abilities. And so we're fostering a whole ICV community to enable that. And again, the good news is these are generalized APIs. Like if they write to IRPIs, it doesn't make it proprietary to HP. We're just trying to build the ecosystem. We believe if we have a vibrant open ecosystem, it will benefit everybody. So if you're writing something that works in our ecosystem, it's going to work on our open stack as well. We think if the ecosystem is very vibrant, it will help everybody. And that's why we're working very closely with folks like Intel, WinRiver, advance of other folks to sort of build this sort of open ecosystem that allows this sort of much more innovation to happen in these apps so that smaller companies, a 20-person company, a 50-person company can build very interesting apps in this space. And that ultimately will help the providers get a lot more innovation in their system. So serious, you guys are really serious about this big business. How big is the market opportunity? What's the TAM? I, again, so there's lots, it's tens of billions of dollars. There's different ways of cutting it. Because the good news from an HP perspective is that every element that this market, well, 90% of this market is all greenfield for us. First of all, that's always good. That's always good. You're always, it's always better to be the disrupted and disrupted. And I've been disrupted. So that's the first thing. The other good news is that across the board, there are revenue opportunities for us. Because first of all, a lot of proprietary hardware is turning into servers. Well, we make a few servers at HP. And pretty good ones. Then there's the cloud layer. You know, and we make a pretty good open source cloud layer, right? Then there's orchestration. And we have NFE director and orchestration. The app ecosystem, where a lot of the money is going to be there, we actually don't want to do two months, because we want to leave it to our partners. Right, HP is a very strong partnering company. But there's a lot of revenue opportunities. There's obviously a huge services opportunity. And I have thousands of services people in the field that can do this. So it's a, from a monetization perspective, that's a huge opportunity for HP. And we're seeing, I mean, it's growing. I mean, I can't disclose the very specifics, but it's growing like a rocket. And then it's growing like a rocket because the industry is shifting. And so really, you know, when I talk about this was the EC and was Megan Martin, I mean, it's a question of, you know, there's a tornado, we have to cast a tornado. But, you know, when you cast a tornado, it's hard. It's hard. It's hard, people get hurt. It's hard. And so we're trying to move, but, you know, the good news is, again, these things require upfront investment. And, you know, Meg and Martin and Antonio and folks that have been 100% behind, made a big investment even before my time when it started with Bethany. And, you know, it's proven to be correct. Investment has worked out, it's taking off for us. And so we're very, very excited about it. That's awesome. Well, it's great to have you back in the cubes or educating us on all this stuff. And we'd love to extract your knowledge and really appreciate it. Great, great to be here in spite of the fact that I've heard it's locked off. Yeah, he's back, you know, lounging around Palo Alto. Louncing around somewhere. From Santa Cruz right now. Good stuff. All right, keep it right there, everybody. We'll be back with our next guest right after this. This is The Cube. We're live from HV Discover.