 Hi everybody, welcome to day two of IBM Edge, it's the Cube Silicon Angles Production, we're going wall-to-wall two-day coverage here at IBM Edge in Las Vegas. I recommend L.A. Bay, about 5,000 attendees here, vast majority of which are customers, about 47,000 customers, and I'd like to, of course, to be IBM Edge partners. And this is year two of Edge. Last year was in Orlando, it was a smaller conference, less than half the size here, sort of the inaugural kickoff of Edge. And basically what's happening now is IBM is expanding that, they're bringing in more systems content, more conversion infrastructure content, even pieces of software business, and really trying to show across correlation and collaboration, across IBM, the different divisions of packaging up a new vision for data infrastructure. In fact, how much layout of the General Manager of the Storage Division said yesterday, his view of storage really should go away as a term of antiquated terms, outdated, and one that is trying to really change in terms of vision that's set forth in the industry. I'm here with my co-host today, Stu. Stu, of course, is a networking and virtualization cloud expert with Wikibon.org. Stu, you had a chance yesterday to observe what was going on. One of the big themes here that we saw, we've met, of course, with Mr. Pearson and Mr. Pearson yesterday, was conversion infrastructure. They've seen much more of networking content. Give us your quick take on what's going on here at the show and in the industry in terms of conversion infrastructure. Sure, Dave. So, yeah, definitely conversion infrastructure is a big topic. IBM Pure Systems really took IBM beyond kind of a blade server architecture. So when you and I were out there launching April of 2012, we really felt that this was the next generation of kind of compute. So if you look at the latest industry numbers, blade server revenue has started to decline and really cloud-type operations, hyper-scale operations and environments that are built for more dense compute environments or where they're going. I had a chance to talk to really the flex system guys and they said this is their workhorse for the next, you know, 5, 10 years in the architecture because when IBM came out with the blade server back in 2001, it was a great technology, a really high performance. It drove a lot of consolidation. It was a key player in the virtualization trend for the last, you know, dozen years or so, but Pure Systems is that new architecture that's going to drive them forward. And, you know, we've been watching all the forces in conversion infrastructure. The other thing that's really important at the show is IBM have a lot of the MSP, what they call the managed service providers are here. These are the people that are going to help bring IBM to more cloud environments. Of course, I made the big acquisition of SoftLayer last week and, you know, looking at how Amazon's changing things with cloud. IBM, of course, is heavily in open stack. Many of you talked to Ambush about that. So, now they've got the open stack. They've got SoftLayer because of the cloud stack. They've got service providers. So, looking at all of these transformational technologies that are helping to go beyond kind of traditional IT as we know. Well, you know, Stu, you've got basically the end points of the vice, really. You've got the dupe on one end and you've got the cloud and the public cloud on the other end, which is absolutely disrupting traditional businesses. You know, generally in storage specifically. Ambush Goyal and IBM set forth yesterday is what I would call a vision. Storage is a platform. And essentially, you know, that is IBM's software-to-five-storage strategy we heard last month from BMC and BMC World and now Viper. IBM is laying down the gauntlet. Ambush Goyal basically without mentioning BMC, trying to actually DMC Viper as sort of SDS, one and a half, maybe even two dot-oh, maybe even one dot-oh, if I had to ask him directly. But really try to put forth a vision that IBM is moving toward the next generation of software-to-five-storage. You know, good vendor positioning, but let's break that down and unpack that a little bit. So, you know, we need to basically get to software-to-five-storage. So, the first one is obvious. You've got to abstract the underlying hardware, right? I mean, and that's something that is really table-staged. So, IBM does it with CSI-Goyal controller, NetApp does it, by the way, NetApp does it. Dave, Dave, I'd like to talk about it for a second. So, you know, SVC's been around for a while, and it is without a doubt one of the leaders in storage virtualization. But really, there's got to be a difference between really virtualization and kind of a software-to-five environment. So, you know, how is this really automating the environment? If I look at what SVC does, you know, traditionally it's really a bunch of, you know, IBM, you know, disks behind it, and it virtualizes that. So, I think that's really, again, so there's a couple things you need for software-to-five-storage. One is virtualizing the underlying hardware infrastructure. But that's not, that's really, it's not even table-stakes, it's kind of fundamental. IBM does that with SVC, NetApp does that inherently. You know, guys like Threepar obviously do that. Etachi obviously has some. EMC does it, I guess, through Vplex. So, that's sort of, okay, storage virtualization one-on-one, but then beyond that, you need to have another layer of abstraction. That looks increasingly, it's becoming open stack or other platforms, VMware, cloud stack, et cetera. And then APIs into that back-end storage layer that allows you to access granular storage services through a RESTful API. And that's where most software-to-five systems break down. So, the question is that you should ask as a customer, first of all, how do you extract that hardware layer? You know, how do you do it? You do it with SVC, you do it with an inherent component of the system like NetApp does, you know, and what overheads does that bring? The second thing you want to ask as a customer is, can I go into through an API and provision storage, not just capacity, but also performance? So, in other words, can I ask for, through an API call 50 gigabytes of storage and 1,000 IOs and then make that my policy and then change that policy on the fly? And the third thing is, what is the nature of that API? Is it a full-blown API? Is it robust? Can I actually make calls to granular storage services and fully exploit it from a higher-level platform like OpenStack? So, that's kind of what I'm taking away from this discussion. What would you add to that? Yeah, Dave, and I totally agree. I know you had an interview yesterday with the group that does what they call the software-defined environments. And what I'd add to that, of course, is IBM plays it across the stack. So, IBM has, you know, lots of things they're pushing in the network and the software-defined network environment and, of course, really pure systems that this server and compute side is kind of filling up a complete stack. And IBM does have, you know, a strong software portfolio because it's really that management, that automation layer that needs to be able to pull all of these pieces together. Tivoli has a strong story. I did talk to some of IBM's partners on the pure systems and they said that the FSM management layer does require a little bit more time to mature. But, you know, at least it's going well in beta testings and definitely some customers have been kicking the tires and looking at this architecture. Now, one of the things that IBM is trying to make as a criterion of software-defined storage is open source. Now, I don't necessarily think that's a criterion of that definition. I do, however, think it's a major advantage for any company that can participate in open source. Of course, IBM is very active in open stack. IBM has a lot of street credit open sources that's made a lot of investments and a lot of money around open source. And so, when you look, actually, that I wrote today on Wiki, when you look at some of the, what's happening within Grizzly and within Cinder, these are the open stack, you know, Cinder is the most important one. Dave, interesting point. You could see that the contributors are, of course, HP, IBM. Interestingly, SolidFire is a huge contributor. Red Hat, obviously, is a huge contributor. Go ahead, Steve. Yeah, I was just going to say, we know that IBM has contributed to open source for a while. They were one of the early drivers of Linux. And, of course, open source, in general, tends to pull dollars away from licenses, so it's away from the Microsoft, the Oracle, and the VMs of the world, and it transfers those dollars typically to services, which is a sweet spot for IBM. So, you know, IBM is great to be open, but there is that underlying way that IBM makes good money off of putting together the solutions that are open. Well, to me, the significance of this is, essentially, as Amazon turned the data center into an API, IBM is essentially trying to turn the storage platform into an API and allow ISVs to actually add value on top of the royale yesterday gave an example of Actifio, and we know Actifio bringing things like, you know, data protection services to the marketplace, being able to exploit IBM's services from, you know, through the SVC. So, a very powerful model at least point forth. To me, the reason this is so interesting is the market leader of VMCs, it's been a consumer internally, you know, used to work there, an internal consumer of open source, but never been a purveyor of open source software. And while it does contribute to OpenStack, it's not one of the leading contributors, but I think that will change, and I think the race is going to be on to see who can leverage that open source better. Okay, for the best. Now, today we've got a full day here. We're going to start off with some customers. Adrian Ledger is here. He's the director of IT infrastructure and SideQuest. And then Karim Abdullah of Sprint is going to be here. We're going to unpack some of the things that customers are doing with IBM technology. Generally, specifically, we're going to learn about flash. And then we've got a big, big, big day today. A lot of server infrastructure, a lot of converged infrastructure, a lot of storage talk, a lot of flash talk. And so, keep it right there. This is SiliconANGLE's Cube. We're live from IBM Edge. And we'll be right back for that for this.