 Live from Las Vegas, Nevada, it's theCUBE at IBM Edge 2014. Brought to you by IBM. Now here are your hosts, Dave Vellante and Stu Miniman. Welcome back to Las Vegas, everybody. We're here at IBM Edge. My name is Dave Vellante. I'm here with my co-host Stu Miniman. John Furrier will be here tomorrow. Jeff Frick will also be co-hosting today. And we're going to talk a little bit about networking. Rohit Mara is here. He's the Vice President of Network Infrastructure at IDC. IDC, of course, the research analyst firm, great reputation, global coverage, does the market sizing, follows all the trends. Rohit, welcome to theCUBE. Great to have you. Thank you so much. So how would you describe where we're at in the networking business? Talk about the changes that have taken place in the last 10 years. And maybe when he talks about cloud, mobile, social, big data, obviously the network infrastructure has to evolve to accommodate these mega trends. How would you describe the changes that have taken place in networking over the last even five years? We are actually at a once in a generation transformation and a complete shift in what's happening in networking. The way I look at it, the networking industry has done well in terms of communications and in terms of bringing voice and data to millions and millions of enterprise users and serving the needs of enterprise IT and network managers for the well over two decades. But where we are right now at the crossroads, as you rightly mentioned, mobile, social, what's happening with cloud and more specifically data center, the velocity behind that, it is certainly a once in a generation shift. And the whole paradigm of how networks have enabled enterprise IT is about to shift or is shifting as we speak. So, I mean, I'm an old guy, I've been around for a long time, I remember S&A, I remember Decknet, and so when the industry exploded, the PC revolution, you saw companies emerge and obviously Cisco's ascendancy, I want to ask you your thoughts on this. Dominant player, now everybody wants a piece of their hide, how was it that Cisco was able to maintain such a dominant share and will they be able to keep it given the disruptions that are coming around software defined and other new technologies that are coming out in the marketplace? What's your take there? Well, you know, no networking discussion of course can be complete without having a discussion around where Cisco's heritage is and where they are going. Certainly they are under pressure in terms of what they bring to the table, how their solutions have been deployed traditionally and what this new age enterprise and network IT managers are looking for and what they are looking for and I think you guys more than anybody else knows this very well is the flexibility and agility that networks really are being required to deliver. Applications are moving at a rapid pace, big data requirements are increasing very, very rapidly. So when it comes to data center networks, the needs are here and now. You cannot be thinking, okay, five years out. These needs are being, I mean, they are required today and data center folks, the IT folks that we speak to, whether it's cloud providers on one hand or it's enterprise network managers on the other hand, they're all looking for that increased level of agility, flexibility, but also keep in mind they do not want to give up. There's reliability and scalability, they have been accustomed to over the last two decades. So to that extent, Cisco has served its purpose in terms of building really world-class, scalable networks that have met the needs of most enterprises and even cloud providers. I mean, there's no question about that if you look at where the market is today and where in terms of Cisco's presence, but there are others in the market who are vying for attention and there are other players, whether it's startups on one hand or other established network infrastructure players, Brocade comes to mind, no question. They are actually number two in the data center networking landscape, Juniper, another major player, and there are others as well, Avaya, and a few others, I can't name all of them here, but for them, this is an opportunity to be disruptive. And I mentioned startups too, that's not the start-ups, right? Absolutely, start-ups are being funded by the Vazoo when it comes to VCs because they see this as an opportunity and I'm coming back to the same thing, the once in a generation kind of a shift, right? So it is an opportunity for startups, for established players, but I also see this as an opportunity for the top guns in networking. The vendors who are there and you mentioned one of them, right? It's an opportunity for themselves, for them to come forth, reassert themselves, meet the needs of what enterprise IT is looking for in terms of agility, flexibility, and I think the other words that I would use, I know they are buzz words, but there's a reason why they are buzz words and that's more in terms of being open and allowing for automation. I think if you can bring the essence of these concepts together, there is an opportunity for vendors to really take themselves to the next level in networking and opportunity they have not had for many, many years. Yeah, Rohit, so I want to unpack that a little bit with you. So if we look at, you know, say the last kind of five to eight years, there was actually a bit of consolidation in the networking world you saw and trying to build a whole stack. So Cisco, of course, launched the UCS, HP bought 3Com, IBM bought BNT, and now with SDN, you're seeing almost it's open for all now. So, you know, since we're here at IBM Edge, I want to, you know, get your opinion as to where IBM fits in because they've actually taken their hardware, the networking group, and pushed that over to Lenovo. Is that because, you know, the trend needs to be more open and, you know, driven by software and they don't need the hardware anymore? You know, where would you put, you know, kind of IBM and the ecosystem and that whole trend of openness? Well, great questions too. I would say IBM's Blade Network portfolio, which was acquired, I think, about four, something like about four or five years ago. It met IBM's need up to a point, but at the same time, it did remain a niche portfolio when it comes to data center networking. It was not IBM's answer to data center networks, you know, as a whole and did not meet the needs of each and every piece of the networking puzzle within the data center, especially when it came to, you know, the integration of infrastructure stack with compute storage and networking playing along, you know, with Blade. Yes, they did have pieces of the puzzle there. So from that perspective, what's very clear is with that piece going away to Lenovo, I think the message that I read from this is that IBM does not as much want to play, at least for now, in the hardware stack, but certainly wants to be part of what they are calling the software-defined environment. So in some ways, I see them taking a disaggregated view of hardware versus software, whether it's through software or through IBM's, you know, other portfolio, whether it's storage, compute or networking, I do see them continuing to make a play in this broader software-defined environment. So they are going to be in networking for a while, as I see it. Again, things could change, you know, you know how our industry is, things change on the fly, but as I see it, as of today, they certainly want to be part of that ecosystem for the software-defined environment of the future, and that includes networking. Yeah, so, you know, we've seen network and hardware kind of been pushed down a little bit. All of these software solutions that are coming out, you know, what do you see as really kind of the huge opportunity of kind of orchestration and creating, you know, better operational efficiencies in the network? Is that the biggest challenge that we're facing? So I see two clear market directions that are going to impact us as an industry over the next, you know, five plus years. And they are both quite tangential. I would say almost anti to each other. One of them is, as you mentioned, you know, the disaggregation of the network, where we're talking, you know, a separate hardware stack. Again, it's a debatable issue in terms of what value hardware adds to the, you know, if you talk to one section of the industry, right from the silicon up, there is a role for hardware to be playing, right? In terms of delivering next generation SDN-enabled networks, right? So from that standpoint, yes, there's some value potentially. If you take another argument, yes, we need to be disaggregating the network. It's a pure overlay play, and you are going to be taking a more layered approach to networking, right, going forward. In any event, it is a disaggregated view which does allow for more open systems, but also allows the network OS to be abstracted from potentially the hardware and allows enterprise and network managers to be thinking more in terms of orchestration and automation. And some of these could be tools that we know of in the industry, and some of these could be purpose-built, but really very kind of proprietary tools that people or cloud providers may want to build for themselves. That's a disaggregated view. On the other hand, though, is the integrated systems view, and I know they go kind of tangential to each other. We see in the industry a big trend to building compute storage and network stacks and putting a management and orchestration wrapper around it. So yes, that's more of a cookie cutter philosophy and an approach, but for enterprise IT, actually there is a growing demand for that kind of agility, that's cookie cutter that you can enable within minutes or seconds. Yeah, networking's full of contradictions these days. We were at the OpenStack Summit last week. I was at the Open Networking Summit a couple of months ago. If you were to ask me a year ago, the whole SDN trend, VMware and Cisco, have been giving a lot of thought leadership. VMware, of course, bought NYSERA who started very much open source and today have almost pulled away a lot from open source and they're working on open interfaces and what they do. As opposed to Cisco has a very hardware-centric view with their whole ACI piece, but now they're embracing open source, they're embracing open stack. Where do you see things like open stack and open daylight and all these open initiatives playing out in this kind of software-led networking? I think they'll play a significant role going forward. No question about that. We see vendors and users in our conversations expressing the clear need that they want to embrace open source. At the same time, there is this concern that if I stay truly open source, will I get that enterprise-grade scalability and reliability that I have been used to over the years? So I think embracing open source is good, but if you follow a model that allows for more rigorous testing, for more validated interoperability across the ecosystem. So whether it's open daylight, I think you mentioned open daylight and open stack, right? There are other examples in that whole open ecosystem. Yeah, there's open Dove and there's open power. Absolutely. As long as there are validated solutions that some of these larger renders can stand behind and certify them and support them. So if you build that whole services organizationally, the services ecosystem around it, I do think open will have a role to play going forward. We're up against the clock, but I wanted to give you the last word. Maybe two things. One is critique IBM's networking strategy. What are they doing and what should they be doing in your view? And last question is what kind of research are you working on? We can close there that is exciting you. Absolutely, and yeah, I know you guys are running against the clock and I need to run from here to a brocade networking and storage session that I'm doing with them in the next hour or so. But when it comes to critiquing them, that's the top one. I am at an IBM event. But certainly happy to critique what IBM is doing. I do think from a networking perspective, I need to see a very clear articulation where IBM needs to go or wants to go from a networking perspective. If IBM is going to play in the integrated stack or the integrated infrastructure need of the data center, hardware or software doesn't matter, there needs to be a clearly spelt out strategy followed by a clear announcement in terms of what are the tactical steps that we're going to see. And even if it's from a software-defined environment perspective, where does networking fit in? And we just talked about OpenDub, for example. Where does Dub fit in? What is your strategy? I think articulating that is the first step. And then obviously I want to see proof points as to how storage and compute and networking will work together. And certainly we cannot forget automation because IBM's customers are also looking for automation of the data center. In terms of your second question, in terms of what's top of mind, I think from a research perspective these days, it's certainly within the data center, it's all about hardware and software from a networking perspective working together, looking at where software-defined networking comes in, what role will the overlay play and the whole discussion around the overlay along with your physical infrastructure. I know the underlay word gets used quite a bit, but really I would prefer to just call out physical infrastructure. How does that play along with the overlay? I think that's an important piece of our research. We're also just about to publish in the next 30 days or so, revised and updated forecasts for software-defined networking and tangentially speaking for NFV as well. Excellent, Rohit. Thanks very much for coming on theCUBE, but thanks for all the great work you guys do in the forecast that you provide the industry. Really great service, so thank you. Okay, keep it right there, everybody. We'll be back with our next guest. We're live here at IBM Edge, and this is theCUBE, we'll be right back.