 And this is Silicon Angles theCUBE. theCUBE is a live mobile studio. We go to events, we extract the signal from the noise, we bring you the guests that are at the show that really help you understand the trends that are happening in the marketplace and give you a flavor for the event itself. We're here live at HP Barcelona at Discover and it's been a great event, three days wall-to-wall coverage. We're here for the bulk of today as well. We've been covering the gamut of infrastructure, of software, really services, HP's overall strategy. We're going to dig in a little bit to server and connectivity and some of the storage and networking trends that are affecting servers. Today's a big day for servers. We've got Antonio Neary coming on, who's now running the HP server business. Jim Gante is going to be here. But right now, Jerome Rubalon is here. He is with the HP server division and Amiya and Greg Shearer who is with Broadcom. He's the vice president of server and storage strategy at Broadcom and a long-time CUBE guest. Gentlemen, welcome to the CUBE. Thank you. Thank you. Good to see you guys. So, Jerome, let's start with you. What's happening? Give us a flavor of what's happening in the European and Amiya region with HP servers. Just give us the quick update. Right, so we recently done a major refresh of our product line, adopting the latest average processor from Intel and it also made big enhancements across the board especially on the networking and IOS side, leveraging some of the unique Broadcom capabilities. So basically now Broadcom became the standard adaptive chipset technology that we have across the board on any of our platforms, for one gig, 10 gig, 10 gig base T, are now becoming standards for built order, configured to order platforms. Well, that's good news. Yep. It's kind of like that, right? It certainly is. It's key. 100 plus billion dollar company and they do a lot of volume there, so you got to like that. So Greg, Ivy Bridge, what does that mean to the industry? From a strategy, from a trend standpoint, why is that important? Well, Ivy Bridge is really just the next sequential launch from an Intel perspective, but it does provide a tremendous amount more compute capability. So we look at the trend in terms of virtual servers. It really enables virtual servers for a whole other avenue. You have a lot more cores, the speed of the overall system is a lot greater, but in addition to that, something that HP has done, which is really helpful for the industry, is their FlexLOM capability, this whole FLR and FLB capability of being able to really get the same use out of, instead of having a soldered down part, you can have an add-in card, but still manage it just like it was soldered down. So all the sideband, out-of-band management capabilities, being able to do things like, from your console or from a network administration standpoint, be able to look at fan speed, power, but do things like program sand boot, so for diskless operations and things. Those are all the things that you got when you had a soldered down LOM, the FlexLOM capability or the FLR and FLB capability for RAC versus the blade environment. All that capability really plays in so nicely now into the Ivy Bridge launch. You have these incredibly powerful servers and now you have the ability to have any combination of IO that you want. You know, one gig, multiple ports of one gig, 10 gig, we're really excited, because within this launch, we now support all the FlexFabric capabilities, so FCOE, iSCSI as well as all the rich networking capabilities. What that means is, is that somebody can start with, say a four port one gig and move up seamlessly to 10 gig, not lose any of their management capabilities just by unplugging these FlexLOM boards, plug in a new one and they have their choice of 10G base T if they're a RAC kind of environment or SFP plus if they are already a short reach copper or want to use optical connections. So, you know, very important that people don't lose capability when they want to swap out something, you know, to a new speed grade. So, I'm curious as to how you guys work together, because obviously Greg, you guys are, you know, the component supplier. Jerome, HP of course talks to channel partners and of course directly talks to customers. How do requirements for what Greg just described come from the field, you know, back to Broadcom and then back through, you know, the HP channel because Greg, what you just described, a customer's not going to tell you or ask you to do what you just described. A customer's going to have a different set of requirements and use a different language. So, what is that language that the customer uses, Jerome, and how does that translate back into what Broadcom developed? Can you describe that information for us? It's been Broadcom, a very long time partners. We've been working together for quite a bit of time. It's not only working from a product qualification standpoint, but it's through the entire lifecycle of the products. So, it's from, you know, engineering up to, you know, supply, you know, production and end deployment. So, really trying to really work tightly together to deliver, you know, higher quality products with the highest level of integration while at the same time, providing what Greg described, which means a lot of flexibility and choice for customers, not looking them down into any specific architecture. So, we do leverage, you know, Broadcom specific unique features like storage offloads, like the CNAs, like Tengik-based TE and we add also the HP innovation on top so that we can drive, you know, greater benefits to, you know, to our customers. Again, everything is based on industry standards, so that's also extremely important for, you know, for HP, never looking at a customer in anything. And also, we do, what we're doing with the FlexSlam is really bringing a lot of intelligence closer to where the applications run, switching all the servers, making sure the service becomes more self-sufficient. And that's definitely, you know, something that we are doing together. So, the customer might say, hey, I don't want to get locked into a specific server box just because of the connectivity in that box. I want to have more flexibility. Okay. Yeah, some of our competitors will push customers, for instance, to adopt, you know, Faber-Channel or Ethernet, but not every customer is ready for that, right? So, we need to give them the flexibility, you know, that they need to adopt, you know, the IO technology that they can. And that's definitely something that we can deliver with Broadcom. So, in thinking about this latest development, how long did it take? What was the development cycle like? You know, it's interesting, because everybody says we're customer-focused, but in your world, Greg, a lot of times you have to sort of decode what the customer wants and put technology out there and, you know, hope you got it right. So, we do, but you know, one of the things that's so great, that you know, I'll turn the clock backwards, is once upon a time, Broadcom's focus was really working with the internal development teams exclusively, you know, at HP, which has been a terrific relationship, but we'd find out from time to time, we'd miss things, because frankly, Jerome and his team, you know, they're touching the customers, they're getting direct feedback from the customers, and it's sort of like, you know, that old game telephone, you know, where when you pass through three or four different folks, you end up- Yeah, it gets lost. It gets diluted. It gets diluted. So, you know- Even in our world, even when you direct connect, things get lost. You interpret things differently, right? Absolutely. It takes years of experience in trying to understand that. So here, one of the great parts is, is we not only have the deep relationship with ISS, you know, in the U.S. with the development teams and the engineering teams, to look at all their specifications and requirements, but you know, we've been really fortunate that we have a great working relationship with Jerome and his team, because even some of the AMIA needs are a little bit different than the North American needs. And to be able to collect those up directly, you know, we actually are engaged now in all the AMIA events. Jerome and his team have terrific coverage across AMIA, the ETSS events that occur. We've started participating in those and not only touching directly the sales folks and the sales architects, but also the end users directly, and resellers and bars. And so, as a matter of fact, the ETSS events, there'll be two of them this year and they're both here in Barcelona. I don't know if they're here in this facility, but they are. What are those, the DT? ETSS is the Expert One Technology Solutions Summit. So it's like our largest, you know, conference for, you know, pre-sales organization, both internally to each people, but also our channel resellers. So it's, you know, thousands of people that are coming just to be, you know, trained to get up to speed up all the latest technologies. And both come as, you know, one of our major sponsor. Joe, Greg was mentioning that the requirements in Europe are somewhat different in the US. Now, I don't know if you can comment on the differences because, you know, maybe you're not in the US, but maybe you could sort of summarize those requirements and maybe Greg, you can help us sort of squint through. What's different between North America and Europe in terms of those needs? Well, I think they're all key, key major trends that we see across the board, right? But I would say in Europe, we are very fragmented type of markets. So, you know, different countries will have different type of requirements. You know, if you take countries like Spain, we'll be very, you know, small and medium businesses orientated, you know, far less corporations. So they really requires a lot of simplification, you know, much more than, you know, larger corporations. And that will change, you know, again from a country to another. So again, very country specific requirements from, you know, specific customers, all of them impacted by the big trends that we see like cloud, mobility, security, big data, you know, that's something that we also see, you know, across the board, across the region. I think then it's more cultural, you know, business type of structure that is, will be different from a country to another. But Dave, something that we've learned, and this has been, I think, a very valuable thing that we bring to the market together, there are several technologies. You know, we talked about FlexFabric, Flex10. This is a technology that they're really HP invented in terms of partitioning, you know, a fat pipe and a multiple smaller, you know, pipes that can have quality of service associated with them for either layer two networking traffic or storage oriented traffic. And that's a pipe utilization, efficiency play. Right, it really is. And mapping back into the hypervisors and native OSes in a way that they can understand it. In this case, it just looks like PCIe physical functions that map back in the OS is really unaware that this partitioning is even happening, that these different virtual devices happen to be sharing the same physical pipe. But something that's pretty unique, that there are other vendors out there in the world that have this kind of technology, but there's only one. And that, you know, HP really stepped forward and said, you know what, this is great technology, but we want to standardize this. We want to make sure that this is available through, you know, industry standard mechanisms. So together, we've taken this into IEEE. So there's a working group in IEEE to take this kind of partitioning capability and allow it to be used across the board. And I think that's sort of testament to HP's courage of- When you say across the board, you mean by competitors. Absolutely. And it really is a testimony towards, you know, you measure something, you know, you put your money where your mouth is. HP's technology is very good. HP has this view of staying ahead. And, but then in addition to, you know, staying ahead of the market, it's giving away the technology so that there isn't the vendor lock-in. Because I think there is a tendency with some of the other big, say, switch vendors that have proprietary environments that there is a lock-in. If you buy into their servers and their switch environments, you can't go anywhere. And I think HP's commitment is is to stay ahead of the market, but make sure to turn that capability over to the standards bodies. And so we have a whole nother section of our groups that work together in, say, the IEEE environments that work in an open industry standard mechanism to go ahead and enable that technology across the board to other competitors. How does that resonate with customers, Jerome? What do you hear from customers when you talk about that sort of openness? HP emphasizes that a lot. What's the customer feedback to that? I mean, I think, you know, customers are extremely cautious on not being locked in any type of infrastructure, right? Could be storage network or servers. They're looking for tight integration, which is extremely important for them. And integrated systems, as an example, is something that is currently booming also in your app. So more and more customers want to see those type of appliance, but it's important that they're not locked in any type of environment. So as Greg mentioned, we do leverage industry standards, but what we also want to deliver is, you know, kill those silos that you have in today's data centers, driving more simplifications, making sure that different IT organizations can work together, so networking, storage, and servers. So there is a big focus within our conversion infrastructure strategy to deliver greater simplification. So if you think like, you know, software defined network on the networking side, you know, direct connect to three power on the storage, you know, blitz system optimizations that we have in terms of, you know, managing the infrastructure. So we're putting a lot of efforts in, you know, streamlining operations, streamlining the hardware architecture that is required, and all of that is always based on industry standards. Okay, so we want to make sure that customers retain their priority to choose, you know, what is the roadmap, okay? So it's hard, right? Because all things being equal, of course people don't want to get locked in, they'd prefer open, but sometimes committees slow you down. So how is it that you've been able to be successful at providing functionality? Because most customers would say, look, if I can't get the function, I'm willing to risk the lock in. So that's a challenge for you. If you want to go open, you want to take it to committees, open source obviously has this. We're not talking about open source, but there's a similar analog here. How is it that you've been able to successfully drive functional best of breed and still maintain that openness? That's a tough balance, isn't it? Right, I mean, if you look at, for instance, you know, software defined network, which is the thing that everybody speaks about today, we're not new in that market, okay? We have been working on SDN for, you know, since 2007, we went through, you know, multiple steps, multiple innovations, multiple proof points on that technology. Right now, we have a complete SDN ecosystem. We have the largest open flow, you know, switching portfolio. We have, you know, SDN controller. We propose SDN, you know, developer kits, you know, to develop your applications again, you know, software defined network. We have an SDN app store, yeah? So step by step, we've built a complete, you know, complete ecosystem, and these, our solutions can manage, you know, a standard legacy environment, as well as an SDN environment at the same time. So we, it's an easy path for the customers to go in that direction. So that's a typical example on how we could drive, you know, industry standards, innovations on top, not looking at customers into it, and you know, making sure that customers can smoothly transition into those next generation technologies. So we're here in the exhibit hall, I think it's hall five, enormous setup in Barcelona. Kind of reminds me of the CBIT in a way, without the zoo, but in terms of this physical space. And we're here at the sort of converged infrastructure and the storage pavilion, and I see all these three-par yellow boxes around. It's interesting, you know, most HP branding isn't yellow, but they stand out in the data center, so I guess they let it go. But so you guys have been announcing converged storage, you know, three-par's doing really well in the market. What, is there anything unique about three-par that you guys can talk about that you're seeing in the marketplace, or Greg that Broadcom is participating in and taking advantage of? Yeah, well I'll give my observations, then I'd love to turn it over to Jerome. But I think one of the things that's so important about this is that, if you look at most storage arrays, most storage companies, they're fairly low-volume companies, so they turn over their server SKUs very infrequently because they just don't have a huge amount of volume. I just look at it sort of from the outside and think, wow, this is a wonderful marriage because essentially all the yellow tags that you see, those are all running on ProLiant-based servers, and so they're really leveraging high-volume, you know, Intel, you know, state-of-the-art servers, so Gen 8, you know, moving to Gen 9, you know, next year and beyond, and you know, looking at that and you think many of the other storage companies can't match that because they just don't have the volume to leverage. So I think that's a really big issue there. You know, of course, in this Gen 8 launch, we now support Broadcom, support FlexFabric, so FCOE capability, and 3PAR has some really, really very nice capabilities in that 3PAR can support a direct connect and supposed to having to have a whole fiber-channel services layer provided by one of the incumbent, you know, native fiber-channel switch vendors that can tend to be pretty heavy for, you know, a small-to-medium business or a small-sand installation where you don't already have that in place, and I think that's a huge benefit. Yeah, we call it the flat-sand capabilities where in fact you can connect, you know, your blade involvement, leveraging of virtual connect technology directly into your 3PAR, and obviously we can now leverage, you know, what come capabilities on the server side, and it's an end-to-end solution where you connect the servers almost directly to that 3PAR array. You basically reduce significantly the cost because you don't need fiber-channel infrastructure in between, reduce latency and speed up, you know, storage deployments. It's a lot less complex when I have a big-sand switch infrastructure. Right. All right, gentlemen, we're out of time, but thanks very much for coming in and sharing some of the server and storage and networking trends, the specifics that are going on in Europe and in EMEA. Really appreciate your time. Yep, thank you. Thank you. Everybody, we're back with our next guest, myself and John Furrier. We're live. This is theCUBE from HP Discover Barcelona. We'll be right back.