 From Las Vegas, it's theCUBE. Covering VMworld 2018. Brought to you by VMware and its ecosystem partners. Hello everyone, welcome back to theCUBE's exclusive coverage, live coverage here in Las Vegas for VMworld 2018. I'm John Furrier with Dave Vellante, co-host of theCUBE, our next guest, Ed Walsh, Cube alumni. He's also the general manager of storage and software defined infrastructure for IBM. Great to see you. Yeah, likewise. So we just, we're talking to Michael Dell earlier and again, an open ecosystem of VMworld again continues to be a little bit more better vibes this year. There seems to be an uplift around, hey, on-premises is working. Cloud seems to have some visibility. It's not gloom and doom for all this data center, but certainly some work going on. You guys were out front on software defined infrastructure and storage from day one. What's the update? What's going on? Well, I think the upbeat here that we're seeing from VMworld is the same we're seeing across the industry. There is almost a renaissance or a boom happening and being data driven, which plays a story heavily. You do have to have right infrastructure to pull this off, either on-premises or in a hybrid or a multi-cloud. Actually, I like how VMware kind of clip, the nomenclature around hybrid is homogeneous across the cloud and multi-cloud is all different types, which we see both. So I think it's helping our business. You're seeing us grow since I've been here, growing pretty consistently above market. You're seeing a large explosion and what we're doing on Flash, mostly because the performance does matter. We did a major new and on. So you're seeing our innovation pipelines alive and really kicking in and it really plays what clients are trying to do. Really be more data-driven. What are some of the drivers of your growth? Some of your successes? What are the key drivers? Yeah, so I think it has to come down to offerings. So if you don't have the right offerings to help people to get, we use the term journey to cloud, but it's really turning more a journey to being data-driven and it has to be in a cloud context. And that is all about having the right offerings. We do focus in different areas for innovation. What we're doing on Flash and it's not just Flash for individual rays, but where that's going as far as how you're going to access the data. It's very different to modernizing a data center compared to what you're going to do on AI infrastructure, cognitive infrastructure, or what you're going to do on these new cloud apps. But they need different things, but it's all about being more data-driven, being more agile, but also really taking the complexity down. So I think you see VMware doing the exact same thing. All the announcements this week were all about making it easier to be data-driven regardless of where you're at, which is what we're seeing. And security too, it continues to pound the security message. At the core, at the core. Which is, you're going to see the same thing from IBM. We're going to talk about technology innovation. We're talking about the industry context, but it has to be the trust and security at the core. And that's where, I think that's why IBM's been doing so well. You know, it's interesting when I've been in this business a long time, you've been in the storage business a while, me a little longer than you guys. But John and I were talking about this the other day. The storage, everybody's been saying infrastructure's dead. I've been hearing that for 35 years. Tom Rosemilius's infrastructure matters. But it's consistently a growth market and a 60% plus gross margin product. So many people don't understand that I hear people that, well, how come the VC's are putting so much money into storage? That company's now, you know, data prediction booming. A lot of companies say, well, we're going to, you know, divest of this or that or get out. You know, you've seen some companies say, well, I'm going to go here. It's just a consistent performer. Your thoughts? So I agree. And it's not just storage. I would say, well, because storage plays into that, right? It's all about being data driven and eventually it has to land someplace for performance. But it's the agility pieces that you're seeing us and the other industry leaders really drive that you can actually leverage this data in different ways. You mentioned backup. So we'll talk about storage rays need to be flexible and fast, but backup ends up really turning into an interesting segment, which you're seeing us lead in, is how do you get more effective? All those copies of data, instead of being some, you know, insurance policy you hope you never get to and it might take you a long time to get the data back. Now these instantaneous copies for recovery, but also for how do you leverage that for being data driven? So also, if you're trying to be doing digital transformation, you have to deal with the infrastructure on-prem or in the cloud. You're looking for ways to do that easier. We used to go out and journey the cloud. It seems more and more it's a journey to being data driven, but it has to be in cloud context. But it plays in a storage or I'll play, you need the right infrastructure to be able to be on this journey. The data driven thing, and you're pounding that message hard, which I love by the way, it's not new for you guys, but one of the things we've been talking about from day one, you know our rap, data is at the center of the value proposition, but we were talking early on years ago about data being a real part of the development process now apps. So one of the things Gelsinger kind of talked about is apps are now the new networks or something around networking, networks. And he also said security, NSX is putting security up around the apps, decoupling that from say network security. So you start to see data in the app component highly coupled with the application, new models around how that's stored and retrieved, service meshes and things going on around the application. How has that changed or have you been vectoring to that place or IBM? So one of our key investment themes is how to do that. And there are a couple of different reasons. How do you become more data driven? So all that data, how do you use it to get better insights, right? And then all the data is typically on, you know, in your infrastructure, a lot of it's on-prem or in the cloud. So one is getting more insights, right? So you need to flexibly go after those copies. And then the other thing is how do you use it to do better app development? So you want to do DevOps, which is, you know, how do you just better quality, faster delivery for a pipeline? But a lot of times if you can't bring the data, bringing like, even if you're doing a test of a mobile app, you need to test it with a production data set. So how do you do that flexibly? And data is the hardest part of all this stuff. So we're doing on AI, how do you get, you know, application driven to go after it in the right performance? But we see DevOps, you can see across our entire product line, an API layer. Yes, we have APIs across all of our storage products and software, but a separate API layer to allow you to do a lot of these things with not to replace any DevOps tools, but to enable those DevOps tools or Kubernetes or whatever orchestration engine to drive the flexibility of your, what I'll say mission critical, either primary storage or secondary storage. Because- So programmability is critical through the APIs. You know, infrastructure is code. If you don't have the right API layer, and we put that in place two years ago. You guys are making that programmable. That's a fact. Okay. And it's in place across our entire portfolio today. So you don't have to have an API for this device or that device or that backup. No, it's an API layer that covers them all. It poises you guys perfectly for the growth of containers, Kubernetes and service measures. So as DevOps starts to move up the stack, you need to have an under the hood programmable model. And that's the software defined. So you guys are saying you have that. Today. So let's- We had you major investments to get there and also do it in such a way that it's a consistent API layer that is abstracted away from the infrastructure. You never want to give, well, in general you don't want good developers access to lower end technology and your system of record data. But you do need to give them access and have the right security around it, access to control. But it has to be API driven. Because sometimes it's not a human. It's DevOps tools. Okay, so let's stay on that for a second if we may. Because Ed, you're known as a business whiz kid. Uh-oh. But people don't realize how technical you are, which probably drives a lot of the people who work for you crazy. But when you think about the ascendancy of virtualization, it changed the way in which, for example, you had to do data protection. And then one of your old companies, you know, you've popularized source side, which made a lot of sense because you were taking 10 servers down to one. As you go into this world of cloud and multi-cloud, you were just touching upon architecturally some of the things you have to do with microservices, et cetera. What are the key components sort of architecturally that you've been driving in your R&D pipeline, which we've noticed over the last two years that you've accelerated at IBM? So we talk about this data-driven multi-cloud architecture. And the only reason for it, it's not, how do you go across a very broad portfolio that IBM Storage has? And how do you have a way to say, we're going to be able to give you a modern, agile, and flexible infrastructure that allows you to participate in modernizing your existing environment or allow you to do AI, you know, literally industry leading type of AI technologies or these new apps. But how do you give clients a way to say this portfolio allows you to do that across your entire portfolio? So one is its API layers. You need to not only have rich APIs around the storage self, but a lot of times you don't want to give those to developers or other people. So we need a separate API layer, which we put across both our primary storage and secondary storage. That was one. That gives you the agility to do almost anything. And it doesn't compete with all these orchestration or DevOps tools. It enables them. It's at the last mile, if you would. Second thing, you need to be software defined. That gives you a way to API, but also literally be able to move things, a flexibility, but also investment protection. And then you need some core innovation, right? So we still make it a lot of hardware. So we're making flash technologies that keep the low latency at workload with encryption on the card at line speed, with Ddupe, et cetera, all the things on the card. So you'll see us innovating on the technology side, but it's also having the agility and flexibility. So you'll see that as a theme. And we can have it in, we see clients adopting it in three areas. It's either modernizing traditional, I would say this is all about modernizing your traditional application infrastructure and how to make it faster, more agile, more private cloud or multi-cloud. The other thing is AI, and it is huge right now getting insights, right? So how do you do machine learning, deep learning, true AI on-prem or in the cloud? And that's different technologies to get that insight. So you saw what we did. The largest AI supercomputer in the world was designed to use 100% IBM storage and actually all IBM systems were the power AI environments. That's a great point about this community. It's really an IT footprint kind of app in its operation. So AI, maybe AI ops, but they're not, AI is not a core competency, like in a TensorFlow. So they need simplicity tools to do that. It's all making it easier, right? So, and I think that's what VMware was doing with a lot of their announcements, just making it easier to deploy. But that's where IBM's been doing, driving that pretty aggressively with our software side, our cloud side, but also what we're doing in infrastructure itself. Essentially, we just had an argument, not argument, on discussion on theCUBE. Yesterday with Jim Covellis and I was saying, I think AI is not in core confidence in this community and he had brought up a good point about AI operations. AI is going to automate a lot of things. But I was saying that's under the hood. There's a general purpose AI, TensorFlow is all these cool, that developers want. Those are the guys that want programmable infrastructure. The guys that want AI ops are going to be under the hood, making things work faster. Infrastructure guys. So- Both are very important. Both are very important, but this is different for Sonos. So like, the mechanics fix the engine. Where do you get your AI? Is it sort of homegrown in your division? Is it a little bit from the Watson group? Where's it come from? So, we like our cloud side. So we have obviously Watson, which is very high end of what we do on a cognitive infrastructure. Very different than normal AI or machine learning, deep learning. But we'll use all the open source tools. So what you're seeing, what we're doing in our, we did the latest product launch, our Flash Desktop 9100, which is granted hardware innovation, NVMe through and through. We talk about AI infused. So it allows you to have better service and support. So what are we using? Machine learning, we say deep learning of all the cold home data allows you to do a lot of analysis. Yes, it's an IBM's cloud. Yes, we use some of the things that we're using in Watson, but it's all the tools. You also see in our power AI for systems or what we're doing in ICP. IBM will go out of the way to say, here's all the open source tools you want to use, Mr. Customer, but you're wrestling with all the open source. We'll help you put it together in a tool without limiting it, but allows you to kind of move forward. Instead of wrestling with the technology, let's move forward. You'll be data driven. We'll make it easier to deploy. So you see that in our storage, but in systems and really kind of an IBM theme, if you would. So I got to ask you a question on the customer impact. You know, we cover the Linux foundation heavily, CNCF, cloud native foundation, you know, doing all those Kubernetes stuff. It's been great being there from the beginning. So, but a lot of, I'm notice a pattern. There's a lot of people talking about new players seeing their software defined infrastructure. I'm going to find storage. You mentioned earlier here that you guys made a significant investment. So the question is how do customers have the potential buyers of this journey of going digital, data driven? How do they determine the people who are saying it and actually doing it? What's required? I mean, because that's ultimately they're trying to squint through the noise as Dave says, what, who's got what? So you got multiple years doing this. If someone says, hey, I got them software defined storage. I just launched last week. They might not have the trajectory. So how do customers test who's got the real deal? It's actually a real, you can, I'm looking at the floor across the way. You have to get past the hype, right? So this is where I like being IBM. So why didn't I do the CEO of the next gig or why did I come to IBM? I think, I think storage or infrastructure is a, I use the term big boy, big girl game. I think it's more than just building the next array. It's how to bring all this together and make it easier for clients to live it the way they want. So I think you'll see a lot of these point products people come up and say, I'm the leader at this or a leader at that. And you saw a couple of the storage guys say, I'm the best AI. Look at my benchmark with Nvidia, right? But then you'll look at like an IBM, maybe not as aggressive in the marketing of the infrastructure, but we're building the largest supercures AI in the world. And we can take the exact same technology and give you a quarter rack or half rack or full rack of the exact same technology knowing you can scale it. So sometimes you have to say is, it's not just the point, hey, you kind of got to dig under the, I would say the culture of the- Is there a question you could customer going to ask, like how many years you've been doing it? What's the investment you've made? I mean, how does a customer truly know if someone's really going to be software defined and positioned for that data driven integration, that holistic package? Is there- Well, one would be, you know, so, software defined, you know, the easy thing is, and this is getting a little technical, but just, okay, show me your proposal and tell me how I can run the exact same infrastructure in a cloud of my choice. And you know, either on-prem or not, or on, you know, software, is that we software defined? And there's a lot of hum and hum on it, then you know that not software defined. Now that in itself is not maybe the biggest thing, but who can you partner with that can help you not only, okay, I'm going to get your next array? That's kind of tactical, but who can I partner with to kind of go forward? And you can't partner with 20 different firms. Who can you partner with to kind of get you from where you are to kind of where you want to get to? So I guess my questions are, I would start asking him, how can you help me in a broader scope? And that's where IBM does well. I think the cloud test is a good one. Can you run the storage on-premise and on any cloud seamlessly? What we did on the last launch is Flashless 9100. We also say it's multi-cloud enabled. And what we did is we put all the software in the bundle. We have these validate designs that show you how to exactly do all the major use cases in any cloud and allow you to actually show you test environments. You do this cookbook at works, but also the software, go try it. But it's kind of, how do you make it real compared to put a bunch of clouds in a configuration? But then it's the use cases. So we do a lot with the leaders and laggards. I mean, we're with the leaders, but then we have some people that really struggle to keep up. And some of it is just showing them the way forward and it's, you got to have a broad enough portfolio to get them over time the right location. Otherwise you end up with a point product here and that point product there. And I think that's where no one has enough team to keep up with that. And you also look at the developers and saying, are developers abstracted away and does the storage, does the performance and versatile for developers without even touching it? Back to your infrastructure as code comment. Well, that'll be the next one. So can I run, you know, Ansible? Tell me whatever technology you're Ansible, if you're going to run, you know, whatever workflow you're going to use for your pipeline, will it work for the API as it should, right? Without opening up the whole storage API. You can drive the car, not even look at the engine. That's the key of storage, right? And it has to be that way. And then can I get all the key abilities in the same function? So it's a good question. I'll think about it more. Well, I didn't mean to put you on the spot there. No, it's good. It's a good question. Different people's, how do you tell the pretenders from the players that's ultimately what customers are? And they do bake-offs, when you see bake-offs, we had an auto-nation earlier. Less and less, right? Well, what's the new bake-off? Just run code on it, right? Yeah, that's right. I don't know, I agree. We're all changing. Hey, thanks for coming on real quick. What's your thoughts on VMworld 2018 this year? What's the vibes you're feeling? What's the overall sentiment? What's your view? I think it's overall very positive. You kind of mentioned early on, it feels like an uplifting. I think you're seeing that across infrastructure. Infrastructure does matter. I think this big data economy, you hear a boom and it's going pretty well. So I think it's exciting and I think you feel it among the ecosystem. Any predictions for next year? What we'll see, because we like to roll the predictions from next year. Predictions for 2019, next year at VMworld. What's going to happen? Oh geez, I think more of the same. I think you're going to see even more and more how to be data-driven. Clouds are given during the cloud, but how to be more data-driven and we're going to make it even easier and easier for our clients to do that. And the ecosystem's driving that. And I think you can see more and more of that here. Yeah, he goes, this is rockin'. Thanks for coming on. This is Cube's coverage here live in Las Vegas. I'm John Furrier, Dave Vellante. Stay with us. We're on only day two. We've got three days of wall-to-wall coverage. Two sets, tons of interviews, a lot of great people, a lot of great content, a lot of data. We've got that data we're sharing with you here. We're data-driven on theCUBE. Stay with us. We'll be back after this short break.