 So John, we have Rob Davis, who's the CTO of Q-Logic, and Rob's been on theCUBE before. So why don't we bring Rob in and have a little discussion about what's going on at the event here and talk a little IO. Rob, why don't you come on in and join us and welcome again, Rob Davis, CTO of Q-Logic. Welcome back to theCUBE. Thank you very much. Thanks for having me. Well, thanks to you guys. I want to just say thanks for supporting us here at theCUBE, who is the Q-Logic. We wouldn't be here, so we are in the Q-Logic booth and I'll see the logos behind us speak to that, so thanks for the support. Our pleasure, you're very welcome. So Rob, the last year or so we've been talking about IO, the changes that are going on in IO, how virtualization in cloud are really pressuring, new storage architectures, new networking architectures, the flattening of the network. We want to talk about some of this, but why don't we start with a discussion around Oracle OpenWorld? Oracle, a little different than one of our last events was VMworld. A lot of virtualization talk at VMworld tons, that's the whole show, more here this year than last year. What are you seeing at this event from Q-Logic's perspective and how's it going for you? Well, I think Oracle has their own virtual operating system products and they've brought them to the show and they're providing those to the enterprise right as competition to VMware. And from an IO perspective, we have designed our products around virtualization to be generic, so they'll work with any virtual operating system, whether it's VMware or Oracle's virtual operating system. So what are you seeing out there as the big drivers, the big tectonic shifts that are changing your business? From the Q-Logic perspective, I think what we see is consolidation of the IO technologies around three basic technologies, Ethernet, InfiniBand, and FiberChannel. In the data center, you mainly have FiberChannel on the sand side and Ethernet on the land side, but with FCOE and iSCSI, there's a bleeding over of the storage technologies into the Ethernet world. So I need to ask you, because Larry Ellison basically trashed Ethernet up on his keynote. It's old, it's at park. So I want to get your comments on his trashing of Ethernet. Two, the keynote was about performance. Obviously Oracle's touting performance is the big thing right now. And with hardware speeds and fees, as everyone's talking about, not a lot of vision, obviously big data is a focus, but performance, you mentioned InfiniBand and all these things. This is kind of in the Q-Logic wheelhouse. So one share with us, one your comment on him dissing Ethernet and two, the emphasis of performance and what really is the performance equation in today's marketplace? Well, from Oracle's perspective, they're big on InfiniBand. And from a performance perspective, InfiniBand comes out of the HPC market, which is high performance by name, right? So I think if you try to compare Ethernet to InfiniBand on a raw performance perspective, InfiniBand wins, but Ethernet is, you know, from a market share perspective, huge. And it's pervasive. And I think you can use InfiniBand in high performance systems that are enclosed around an application like Oracle does. But if you're going to connect to the rest of the world, you better have an Ethernet plan. So we were on the Twitter stream, obviously on the keynote, and all day yesterday, and there's been a debate going on around the notion of general purpose computing and special purpose-built computing. Obviously Oracle's going more purpose-built, integration all the way up and down, and then you've got this general purpose where Ethernet does fit into a more open ecosystem with converged networking. Talk about the approaches, and how do you see that playing out, the purpose-built architectures and the general purpose architectures? Well, first of all, we're focused on providing high performance IO to all of the different approaches. So we do InfiniBand technology, Ethernet technology and fiber-tech channel technology, and we're able to mix them together in our products. But as far as where one should be used or the way the industry's going, we see that our OEM customers are trying to get more and more of the solution and more and more of the stack into their product lines. Whereas it used to be, you would go to Cisco for your networking, you would go to EMC for your storage, you would go to HP Dell, IBM for your servers. Now they're trying to put those together into one stack and wrap a certain technology around it that Oracle has chosen in InfiniBand as that technology for the IO. So two questions, what does that mean for you? And then I got a follow-up. For us, it means that we need to make sure that we're providing the technology choice of our OEM. So from Oracle perspective, that's InfiniBand from other OEMs, it's Ethernet, fiber-channel. So my follow-up on that is that virtualization really changes so many things. From a performance, from a monitoring standpoint, in a physical world, you can kind of see what's going on. You know what the connections look like. You can put fences around different physical resources and secure things in a virtual world. It's like one big black box. Are you seeing demand from either your OEM customers or maybe, and I know you're one step removed, but from end customers, you know, big banks and insurance companies and the like to get end-to-end visibility so they can do better management, more performance tuning in a virtualized world? Well, from our product line perspective, we provide the hooks, you know, the dials and the meters for our OEM customers to integrate their management. In the world of Oracle, for their virtual operating system, we have special registers in our adapters that provide the functionality needed for moving the VMs you know, across to different platforms or for tuning the performance around a particular VMs or requirements, those kind of functions. So I want to stay on this topic for a little bit because I'm trying to understand the value of end-to-end if in fact there is any. Is that illusory? Is it sort of just hype? Or do you see that there is value? Well, I think the end users think there's value because if they've got products from three or four different companies, say doing network, doing storage, doing server, they need three different management platforms. They need training on those. If they can get it all from one company like Oracle, then they have one pane of glass, one management tool that manages the storage, the networking and the server as well as the operating system. So we like to look at the horses in the track and you guys have obviously big adapter business. You've got sort of switches in the edge. They're getting bigger. You're really not dominant in the core at this time but you've got a lot of the pieces in place. It's certainly brocade is pushing into end visibility with its strategy. Dell buys force 10. Now they don't have the adapter side of the business but they've got the top of rack switch in the core. I'm sensing that the industry is getting closer to competing on that basis. Do you mean competing on a total end-to-end solution? Oh, totally, totally. What does that mean for you guys? Can you get there, particularly the white space being the core, are your edge switches getting large enough now that you can sort of rack and stack them and deliver that vision? Again, we go through OEMs. So for the perspective of Dell buying force 10, that's very similar to in our announced product lines, HP bought 3Com and we're embedded in 3Com's products that are now HP labeled. So our goal would be to do the same with the acquisition of Dell of force 10 with the acquisition of when IBM bought B&T. We have products that interface into these new networking products of our OEMs. Well, Dell's an announced customer, right? Dell's an announced customer for certain products, yes. For example, I think you've got a design win there. I lost track of the design wins, Rob. I mean, you've got so many of them. Thanks. How about this notion of management? Talk a little bit more about your management philosophy again and how virtualization changes that. So our management philosophy is to allow our OEMs management software to interface in ways with our hardware that gives them value to their customers. My question for you is obviously, you know, Oracle's got the new approach, well, same approach on one end, but with big data and being vertically integrated with the performance, has the OEM business changed for you guys over the past few years? And can you share with the folks out there, you know, what major seismic changes have happened in the business and how you guys approach working with all the top players? Because you do have insight into roadmaps and architectures because you are, you know, dealing with future products. What's changed? Has anything changed? Well, I can't really share the roadmaps of OEMs. Sorry, but what's changed for us is that we used to be able to take a product like a fiber channel adapter back in the days of 2GIG and we would sell that to all the OEMs with hardly any changes. And now as the OEMs build these vertical product lines, we are spending a lot of effort to figure out how they want to differentiate from the other vertical product lines and that means customization to our products different for each OEM. So we have Intel on later and we're really interested in hearing what they have to say. You know, John and I were talking earlier about, you know, the generations of microprocessors that we've seen going from the PC era which Intel clearly dominates and now it looks like, you know, they're dominating the server business and of course you've got new mobile trends going on but Romley is something that the industry is expecting. Coming up here, you know, shortly. What does that mean to you? From our perspective it means generation three PCI Express. So our products that are designed for the Romley timeframe on the adapter side have Gen three interface that in turn gives a higher performance to the wire coming out of the adapter. So for fiber channel, it's our 16 gig fiber channel product that we announced last week. For the switches, it's 40 gig ethernet, you know, so the speeds are just getting faster. On the virtualization side, it's more features for the virtual operating systems in the adapters, so on and so on. I wonder if you could bring out your binoculars or maybe your telescope, maybe go a little bit further. You've been with Q-Logic, how long now? Since the acquisition of Ancor in 2000. 2000, so you've seen, you know, the transformation of the company. Pretty dramatic transformation of the company. Absolutely. When you guys made some big bets on integrating silicon and a lot of people thought you were crazy at the time but you really sort of evolved, you are evolving I would say from what I would consider a pure adapter company to much more of a data center supplier. Is that an accurate perception? And can you talk about what you're trying to do there and what your vision is for the company and take the technical perspective if you like. But I'd like you to share with the audience where you see, you know, the company going, you know, midterm to longterm. So I think our, you know, we're an arm supplier to the OEMs, right? So they want fiber channel technology, we have it. They want ethernet technology, we have it. They want InfiniBand technology, we have it. So our approach is to make sure we're designing products today for an intercept point with them in two or three year time frame. Cause that's the time it takes for ASICs to be developed and debugged. So do you see yourself as a data center technology supplier or more of, we see ourselves as a supplier to the companies that provide, you know, go to the data center. We don't go direct, we go through OEMs, yeah. Right, right. Okay, good. Rob, Davis, thanks very much. You know, appreciate you coming on theCUBE and sharing some of your perspectives about what's different about Oracle, where Q-Logic is headed. And I'm sure we'll see you around the circuit. Thanks very much. Thanks for your time. Thanks for your time. Take care.