 This is theCUBE, we're here live in Silicon Valley. This is SiliconANGLE, Wikibon's theCUBE, our flagship program. We go out to the events, extract the signal from the noise. I'm John Furrier, the founder of SiliconANGLE. I'm John Michael, Dave Alonzo, the co-founder of wikibon.org, and we are excited to be here at the Open Compute Summit, and we have Lakshmi Mandium here, who's with ARM. Thank you for coming on theCUBE. CUBE alumni, we last interviewed you at the Moonshot event in New York City for HP. And Moonshot, great, great architecture. It's a lot like kind of what the mojo is here, which is, hey, let's get the cost down, let's get the performance up. A little bit different, different approach. So welcome back. Thank you, thank you for having me. It's exciting to be here and see the energy level. I think especially watching the keynotes this morning with Frank, and it was really good. Zuckerberg shows. Yeah. It's recent tomorrow. That's something I can tell my teenage son about, and he'd be proud of. I saw the founder of Facebook. The founder of Facebook was here. We were gawking at him. He couldn't come on theCUBE, but we'll get him next time. Were you here last year? I wasn't here last year. No, so John, you were here. I wasn't here either. But compare this year with last year. Oh, 10 times better in terms of size and vibe. Last year, it was just clearly off the runway in terms of, wow, this is really happening. Kind of raw. It rocked the house. It was getting people getting behind it. Now it's to me full on, you know, this is a mandate. People are lining up behind the ecosystem developing. They have booths here and vendors. It's just a much different show. The audience packed house. That looked like a Mac world. It looked like a, you know, a big Google I.O. type event where everything's packed, the whole show is packed. Yeah, I think they doubled their registrants, I think, from last time to this time. At least that's what I heard. It feels like it's just about to explode, right? So, okay, so why is that? You're in the center of all this. I just think people are looking at ways, especially in the data center, right? You heard, you know, I think Cole Crawford said it in the morning. He said, you know, just from IOT alone. In China, there's going to be 100 Zeta bytes worth of data being generated. So you think about that kind of burden on the data center. You can't really do things the way you were doing them before. And I think people are looking at how they can innovate. And I think Open Compute provides a platform for them and a forum for them to be able to talk about innovative ideas. I mean, ARM's been a part of it. We were announced at the last Open Compute Summit last year. We've been participating in it. You know, my team participating on several of the working groups. We have partners like AMD who announced that they're going to have a ARM-based Open Compute contribution today. We've got partners like Applied Micro who did the group hug board. And I think actually this whole thing is a big group hug. I think Frank said it this morning when he said, AMD Applied, followed by Intel, was a group hug kind of. So obviously ARM, you're knocking out of the park in the mobile, you're in the server space. The great thing is you don't have all this baggage in your code base, but you got a lot of work to do. So talk about where we're at. Give us the update on the server business. It's, you know, I think today is a great example, right? We had two partners, Applied Micro, talking about their 64-bit solutions. We had AMD also talking about their 64-bit solutions and how, you know, Applied's sampling their second generation this year and AMD is going to be sampling, and they have boards here. And this is a great showcase of how vendors are able to come to market fairly quickly and with 64-bit solutions, they're going to really address a broad set of applications. We also have one of our partners, Marvell, who has their 32-bit storage server in the We Win booth. And this, of course, was the one that was deployed famously at Baidu. So we're seeing, you know, more and more solutions coming out there in the marketplace. And I think the software ecosystem and the hardware ecosystem are coalescing. You know, there's several, again, We Win as an example of an ODM that's got an ARM-based server solution. People like Mitak have announced solutions. And of course, you've got HP with their Moonshot and Dell has also had several, I think last year Dell showed their group, Hugboard. So, you know, there's quite a bit of momentum building. Dave was kind enough to give me props on his panel. So I'll give Dave props. Dave was hosting the keynote panel this morning with all the luminaries. One of the things he quoted me on was, I said, there's a reset going on in the hardware because the physical packaging needs to get tweaked because there's so much software going on, software to find everything, even bare metal software as Cumulus was arguing. But also this homebrew club where, you know, that's a vibe here, it's more of a cultural thing. People feel like it's the next homebrew kind of tinkering environment where the next breakout's going to happen. But with that, in the homebrew days, the semiconductor industry drove that innovation. So I got to ask you, is there an element like the microprocessor did for Apple Mac, which is celebrated its 30th anniversary? Is there a version today? And is it armed, possibly? What is that semiconductor innovation that is going to catapult this next band of misfit, the crazy ones, the ones that are going to build the next generation? I think Raspberry Pi is a great example of that. There's a whole movement around Raspberry Pi and people doing development and then armed from a, if you look at the next wave of things that are coming like the internet of things, sorry to play a little bit of buzzword bingo, but we actually have a developer system called Embed, which is focused on that next class of developers that are doing these IoT or sensor kind of applications. And so I think there's multiple spheres of homebrew clubs going on already in the embedded world with things like Embedded with Raspberry Pi. And of course, from a software perspective, all the mobile development that's happening, that's all happening on ARM and there are multiple partners that have development kits that are driving that. And then I think you're starting to see that now coming into the data center and the server space with open compute and again, the availability of software development kits from partners like Applied Micro, like AMD is also going to foster that kind of software innovation. It's like the DevOps world, if you look at the cloud, DevOps has always been, well, we don't have an engineering, an ops department, it's just one department called engineering. You're seeing the same thing with hardware and software now. You're seeing hardware geek saying, hey, you know, it's like what DevOps is the cloud, that same trend has happened in Embedded, but now it's going mainstream. Yeah, I mean, and another interesting thing is you talk about how do you optimize the software and hardware worlds together, right? So ARM is bringing things like integration, we're also bringing things like heterogeneous processing. I think last time when we were in New York, we talked about TI with their DSP capability and how bringing that heterogeneous processing environment was going to create more performance per watt per dollar, that whole TCO conversation. And I think that trend is only going to increase. I think today you heard Applied Micro talk about their networking capabilities. So all of those things, you know, you had, actually I tweeted your statistic about converged infrastructure $400 million and data center was, yeah, I'm sorry, billion, yeah, with a B and data center infrastructure was $80 billion. And I think from an ARM perspective, we play in both of those because if, whether you're in a converged infrastructure world where you're wanting to integrate multiple functionalities together, the degrees of freedom that you have with all the innovative choices from the ARM partners is pretty high. So Luxman, I want to ask you how you look at segmenting the server business because in a way you can spray and pray and try to put as many seeds out there as possible. But I would imagine the old sort of price, IDC price bands just is not the right way to segment this market. You're going after sort of somewhat unique use cases, although some very large scale, but how do you look at segmenting the server business? Because the model that you use is 30, 40 years old. How do you look at it? I think it's really going to be on the workload focus. So you can look at things like NFV, right, which is a networking function virtualization which is being driven by mobile operators and SDN, right? Some would argue that those things are starting to come closer and closer towards traditional servers. And then you look at the definition of one P2P doesn't really fit anymore when you look at all these integrated high ultra dense. I mean, I think some people call them ultra dense, but I think over a period of time that you're right, the line is wavering between what's a one P2P, it's just, I don't think, we don't view it as a segmented that way. We view it as what workloads, how do we go after those workloads? What's kind of the overall performance that they're looking for? So I think that's the way things are going to evolve. Might make it easier for some analysts, I mean harder for some analysts to track, but I think from an end user perspective they're going to classify it. Harder from the standpoint of being accurate, maybe easier, because nobody will know the answer. So where do you expect to have traction? Where are you seeing sort of early signs of momentum? So, you know, storage was a great example that I talked about, you know, it's on the floor here where people are seeing the value proposition of integrating, you know, the lower thermal footprint from an SOC using ARM, the integration capabilities. We see things like memcashity, again, more IO intensive workloads, content delivery networks, static web. We also see things, you know, also in the new world where people are doing more scripted languages. For example, there's an organization called Lanaro, a Lanaro enterprise group, where it's a number of companies in the ARM ecosystem that have come together to work on Linux. And they've gone in the three years of their existence, they've gone to becoming the number two contributor of Linux to the kernel. And so in that, there's a group called the Lanaro Enterprise Group, and people like Facebook participate in that and have ported their hip hop virtual framework over to the ARM architecture. So that's where we see, you know, scripted languages or where there is a real IO intensive workload, that's where we see ARM kind of really making a difference in gaining traction. What do you make of the rise of the ODMs? I commented this morning on Matt Eastwood's numbers, about 7% of the revenue, 12, 13% of the actual volume. A lot of people a few years ago, I think were maybe hoping that they would just sort of blip and go away. It doesn't seem to be the case. It looks like they're getting a really strong foothold. You see IBM exiting the x86 market, Lenovo maybe in a much better position in theory anyway to compete with the rising ODMs. What do you make of all the sort of shifting sands of competition in the server business? I see opportunity, because I think the ODMs are also trying to find their place in the world and they want to see a place where they're kind of moving up the value chain in terms of offering more customization, offering more services perhaps and all built on kind of their traditional foundations of efficiency in terms of operations, right? And I think they're looking at things like the ARM ecosystem story to help them differentiate and help them tailor. And actually, if you look at the hyperscale kind of deployments that they are servicing, large scale guys, I think Facebook talked about how they deploy their own infrastructure. There's a real opportunity to tailor to a given workload. And I think this is where the ARM strength is going to come into play in terms of them being able to say, hey, you know what? I know there's 300,000 servers that are being deployed for this particular workload. I'll work with an ARM partner that has a tailored processing solution for that and I can still add value, get efficiency and satisfy an end user. So I think it's opportunity for everyone. Do you think the, I mean, if you look at the server business historically, it's one guy made a lot of money in Intel and everybody else sort of picked up the pieces. I remember Scott McNealy saying, hey, you got car makers and you got car dealers. Car dealers made a lot of dough. They all had nice cars, they lived in big houses but the car makers in Detroit were the ones who were the big barons of industry and of course we know how that worked out for Son. They, you know, valiant try but couldn't make it. With the server trends that we're seeing now, do you see that dynamic changing where there's a lot more, maybe the margins aren't as great, maybe there's not a monopolistic sort of dynamic going on but there are, as you say, some exciting opportunities out there. How was it different today than it was, say, 15 years ago? First of all, I think the rise of open source has opened, no pun intended, but you know, open doors for new architectures. Huge wildcard too, right? Yeah, and I think you see the ARM partnership entering the market has already changed people's perceptions of what's possible and I think you'll see that it's going to drive a faster pace of innovation. We saw what happened in the mobile ecosystem, right? Think about your mobile phone. It's changed over the last five years in terms of what's possible and it's been largely because there's been a huge number of choices and innovation in terms of supply chain, in terms of new IP that's being integrated and I expect to see the same thing happen in the data center space because now you have all of these choices and people are innovating at different paces but it's still overall accelerating the pace of innovation in the market. So John and I actually see a renaissance coming in the data center. You know, a lot of people are concerned, tight margins, prices falling, et cetera. We see that as actually a positive trend for innovation. We think the market is elastic and it'll actually create new growth, new opportunities. Did you buy that premise? I do because I think it, if you think about new businesses that are emerging, you know, like whether it's the knowledge economy that we heard about this morning, right? It's really about how do you lower the barrier of entry for a new business engine, right? And so if you think about people starting the next Facebook as an example, whatever it may be, the barrier to entry historically has been infrastructure and capital costs and I think with this renaissance that we're seeing and the new levels of innovation that we're seeing, you'll see the barrier to entry for new businesses is lower. I think, Dave, one of the things I see that's obvious is arm is proven that the architecture works. Developers are adopting it. What I'm interested in is the enterprise because that's been a market that's been dominated by Intel. And that's what the demand is. When you're looking at like what IO did today was interesting, right? They demoed an iPhone spinning up on a cloud. 600 customers. How does that happen? I mean, this is a company that was hosting years ago. It's now in the same discussion as Google, Amazon, and these big, big cloud providers literally overnight. So to me, I think the economics and the performance changes are going to change the game. So I got to ask you on that thread, what things do you guys see in terms of proof points? If you buy that, the commoditization of innovations happening at the same time and certain people will win, certain people will lose, what proof points do you see that point to that? I mentioned, Aya, what are the ones that you see as great examples of the new school, the new innovators? I mean, I think China, just China. The level of adoption that we see in terms of new business coming online, new hardware suppliers, new semiconductor suppliers across multiple different markets is a great example of where it's really accelerating the pace of innovation. It didn't mean to sound flippant, but you know what I mean. The Internet of Things, it's everything, right? Internet of Things, but it's not just about the thing, the sensor that ARM is predominantly in, but it's also about the services that's going to drive back into the infrastructure. And again, this is where... We're going to do some compute requirements. There are compute requirements. And I can remember, I said the 100 Zeta bytes of data that's going to be generated in China alone. You think about that. How are you going to process that effectively, cost-effectively in a fashion that doesn't cause you to build 50 power plants just to sustain it? And I think that's where you need innovation and choice, and that's where the ARM partnership wants to make a difference and drive that. I got one more question before you break. So I got to ask actually two more questions. So what's the hardest part about what you're doing in terms of trying to crack this market and what's the most fun? You know, hard and fun, I think, are kind of interchangeable in that if it's not hard, it's not fun. And so, you know, for me, it's just seeing people bringing ARM into the conversation or even thinking that there could have been an alternative when five years ago it was like, oh, I just got to take whatever is dished out to me and live with it. I think just the choice and being able to disrupt an industry that has been kind of stale for a little while in terms of the supply base. And for me, it's just changing the world in a way, right? Because you're now bringing new players into a conversation, you're spurring innovation. And for me, that's the fun part, is say, how can I change something? How can I not accept the status quo and change something? You're going in a market where you had one architecture was 98% of the volume or whatever and now bringing a new architecture, a new set of players and a new set of thinking into that market, to me, is what excites me about it. So, all right. Okay, final question for you is, talk about what's the bumper sticker of this show? Describe for the folks, because you're in the semiconductor business, you kind of know the innovations, so it's kind of a trick question to say, what's next? Describe for the folks out there what's happening? Why is Open Compute so important? What's the bumper sticker on the car? What's happening here? Why is it so compelling? Why is it a revolution? I think the bumper sticker is, don't accept the status quo. Don't accept the status quo. Don't settle. Yeah, don't settle, don't accept the status quo, and there's always a choice. It's the Steve Jobs commercial, the crazy ones are the ones that actually make the change happen. A lot of great innovation happening here, it's really exciting, I mean, you guys have done a great job enabling that. I think it's just the beginning, so congratulations. Thank you, I appreciate that. Okay, this is theCUBE, we're right back, we're here with Arm Holdings, making things happen, I mean, everything you touch these days has a arm processor in it pretty much, so these guys are doing great now, and the enterprise continuing to push the innovation. We're right back with our next guest after this short break.