 We're back live at the IBM storage summit. And last week at SuperCloud, we talked a lot about chaos, multi-cloud chaos, cross-cloud, Kubernetes chaos. It's a complex out there. So how do you tack complexity? We're here with Pete Bray, the global product executive at IBM, to talk about just that. Pete, good to see you. Great to see you too. So can you explain the good and the bad of Kubernetes? Oh man, that's a, how long do we have? Yeah. There's a lot of good, but you know, you don't just, what's that tweet? You don't just deploy Kubernetes. Exactly. Yeah, there's a lot of goodness obviously. It delivers the agility that the organizations need, you know. And what's interesting is to see this trend that's been transpiring as organizations, and this is kind of the bad part of it is, we're going through a transition period. And organizations are wrapping their minds around this new technology, this new way of doing things. And it's not just the technology, it's changing their organizations and the roles of the people in the organizations. And you know, I know we talked at KubeCon in Amsterdam, and we touched on this a little bit, but you know, the advent of the platform is becoming very, very critical for these organizations. They're looking for a single, simple solution to be able to solve these problems in this world of complexity that they're dealing with. Yeah, I think that when we were talking at KubeCon, it was really around platform engineering and how that was becoming a thing. And the skill sets though are just far and few between. What are the things that you're doing now at IBM to help kind of bring that together? Yeah, and one of the big changes that has happened since we talked at KubeCon is now these organizations are looking not just for a platform for their cloud native needs and for their Kubernetes apps, but now they're looking, I want to bring my virtual machines over too. And this has taken off like wildfire, you know, and we work very closely obviously with our friends at Red Hat, you know, and they have a technology called OpenShift Virtualization, you know, based on Kube, the OpenShift or the open source product. And we're working very closely with them to understand, you know, the use cases that these customers, you know, bringing these VMs over want to be able to co-host not just the containers, but also the VMs together and have a single substrate to support it all. And that's all on OpenShift on IBM kit, right? Yeah, exactly, exactly. And that's really what Fusion is all about is, you know, we'll support, you know, and again, talking about the complexity on-prem deployment, you know, hyperconverged solution, but also the software solution that you can run anywhere in public cloud, on-prem, bare metal, virtual machines. You know, we give the ultimate and flexibility because we know that these organizations are going through this transition right now and they have more questions than they have answers. Yeah, I mean, that is the, you're talking about on-prem and wherever. It is the enabler, if you will, for that cross-cloud capability. When Kubernetes first came out, it was, you know, relatively immature. You had other sort of container platforms that had these enterprise features, and of course, Kubernetes tried to keep it simple in the committers and it worked. But now, Pete, when you start to talk about all these different estates and supporting different locations, you've got to build in those enterprise recovery capabilities. So what are those, what are the piece parts that you're bringing in to simplify things for customers? That's a really good point. And one of the things that I wanted to talk about today is, you know, this overarching concept of both the applications and the data delivered as a platform. And the ability to provide really to turn the data into a revenue-able asset. You know, I've been reading a lot about your super cloud concept and how that touches on the importance of data and some of the examples that are out there. And it's really, it's not about the data itself, it's about what you do with the data. How do you massage it? How do you make it available? And so we have this concept called the Fusion 5. And it's the persistence, basic capabilities. Can I store and retrieve the data, but do that in a cloud context? It's the resilience. How can I be assured that the data is gonna be available when and where I need it? And I mean, we're all from the storage industry here. So, you know, talking about things like snapshots and backup recovery, disaster recovery, but doing that in a multi-cloud context and delivering those kinds of capabilities. The security, obviously, you know, we talk about cyber resilience, but there's even more fundamental capabilities around encryption and key management and user authentication that the expectation today is that that's just included with the solution. I don't need to bolt something on. And this is the recurring theme we're seeing from the customers that we work with is they want a turnkey solution. We did a lot of research this last year with the customers that we have for Fusion. And that was the one biggest thing that they told us is we value the single platform that IBM delivers. It includes all the OpenShift. It includes all these additional data services. If I want a turnkey hyper-converged system, I can get that too from IBM, but I can also get the software from them. The next area, which is really a key area, is data mobility. The ability to move the data together with the applications across clouds, on-prem, at the edge. This is really a key capability that we're starting to see. And it's not just for re-platforming or re-hosting applications. It's also just for everyday needs, being able to move data between AppDev and AppTest environments into production and back and forth is very critical capability. And then the final area, which is really unique to IBM and a lot of our Watson heritage, and I had to work in the Watson reference, given it's IBM, but data cataloging. Talk about all these other capabilities making it available and secure, but what good is it if you can't find the data at the right time? And so we have unique capabilities in terms of being able to catalog and label and tag the data so it's quickly and easily found. And you got to do this, like you said, across clouds. So you have to have all those storage services and other services that you talked about, whether it's snapshot or encryption, and it's got to work the same. It's got to have that consistent where the super cloud comes in. It's got to have that consistent look, feel, developer environment. And then the data cataloging, explain that piece a little bit more because it's a hot topic these days. That just enables optionality, different data formats, different types of data. Yeah, yeah. So I mean, and there's stats that are floating around out there that the number one problem for the data scientist today is not, how long my inferencing takes or not how long it takes to do model training. It's, can I get to the right data quickly? And some of the estimates are like 80 to 90% of their time and spent just trying to find the right data. And that's the problem that we solve. On ingest, we're able to tag and label the data so that later they can run queries and they can quickly find the right data set at the right time. Yeah, it would seem that again, it's taking it from being a storage platform to being that data platform and really evolving that. Yeah. And that, is that working in concert with a lot of the OpenShift from the Red Hat and what they're doing underneath the hood as well? Yeah, and in fact, I'm gonna steal Dave's, I'm gonna riff off of Dave's concept of SuperCloud a little bit and talk about the super platform. You know, and by all means, I think nobody would argue that OpenShift is, you know, got a big leadership position in the Kubernetes space. And that clearly addresses the application side of the equation. And even now, you know, with this OpenShift virtualization capability, being able to bring in existing VMs, host those, while they're going through their transformation to become cloud native applications, we're also providing the data services that go along with those applications. So now you get both in this super platform, as I call it, that you can run anywhere on any cloud, on-prem, bare metal, you know, whatever you need. So you talked earlier about the skills and the roles. So how does this, what you're doing, what are you hearing from customers in terms of how it affects their ability to actually get to deployment, get to value with less stovepipe skills? Yeah, and it's interesting to see the industry change and individual customers change and make these realizations. They've realized that the traditional skills aren't going to carry them into the next decades. And so they have to change. And the way that they're deciding in going about this is rather than trying to take existing people and retrain them, and this is a consistent theme we hear, you talk about the good, the bad, the ugly of Kubernetes is fundamentally it's a new technology and it requires a new set of skills and retraining. What CIOs have realized is I need a separate organization that addresses this and addresses it more holistically from a platform standpoint, thus the name platform. And we have now platform architects, platform engineers who are charged with building these solutions. And a lot of it's being driven by the challenges that these organizations have, trying to figure out, do I repatriate, do I, you know. You can say it, you can say it. Yeah, do I move applications around? What about my public cloud costs? You know, how do I optimize that entire environment? Kind of leading up to the super cloud concept. They're dealing with that challenge and they've realized I've got to have an organization that is focused on this, that is peaked in skills that really understands these technologies because they are different in many ways. But what's really cool to see though is now this new capability to be able to support these legacy environments alongside these new environments. It helps remove some of that complexity that they have to deal with. And you're definitely seeing in the marketplace, so more of an equilibrium or a balance. You know, the big theme earlier this year was cloud optimization. People trying to sort of reduce their cost. You know, database has been a big driver of cost, not just compute that shows up in the surveys. And I think there's more data in the cloud than a lot of the cloud guys would have you believe. You know, they say, oh, it's only 10% much higher than that, you know, we think. And so as a result, the stuff that should have moved to the cloud, I think most of it or much of it has moved. Okay, I think we can pretty much agree on that. There's a lot of new activity and new innovation going on in the cloud. But a lot of the stuff that shouldn't move to the cloud is probably not going to move now, especially because you get that cloud operating experience anywhere. Right, right, yeah. And I think you'll see this that, I mean, we talk about hybrid cloud, it's real. I mean, it's because of the points that you just made. You know, there are applications that are best you met for on-prem environments. I think the whole super cloud concept, it really is what hybrid cloud and multi-cloud envisioned. And it's now coming to fruition. Of course, you inject AI and it just brings a whole new conversation. What are you hearing from customers in that context? Absolutely, and they are looking, how can I monetize this data asset that I have? And AI is a key integral part of being able to do that. But even like Kubernetes is a brand new technology, AI for a lot of organizations is still very new. And this is the challenge for us at IBM is we have to get outside our own echo chamber sometimes. We're immersed in all these different technologies. And we have to remember that sometimes at the end of the day, the customers, they're not that advanced yet. By and large, there's a lot of customers that are and they grab it very quickly and run with it. But we need to help them. And the way that we can help them is deliver simpler and easier solutions. And one of the points about Kubernetes, the good, is it was a fundamental concept at the beginning, Kubernetes, to if it's automatable, automate it. And that's really, I think, one of the key things that's going to carry us into the future with Kubernetes is the automation that's going to be built in and baked in. Yeah, I think that's a big piece of it for me, is that as I've talked to customers, especially once you get out of the Fortune 500 and into the more global 2000, those skill sets are just not there. And a lot of them, IT has become platform engineering, but they're losing some of the skill sets even on the hardware side. And I would assume that that's what you're hearing about bringing in and being able to bring in whatever kind of solution they need is the easy button concept that we were talking with one of your ecosystem partners earlier. Yeah, and I think that's the exciting thing, being a techie guy, seeing a technology like this go through this evolution and opening it up to a much greater market to help these, the global 2000 like you talk about, to help them be able to achieve the benefits that the big guys are able to achieve. I want to share some data. You're talking about customers are still trying to figure out how to use AI. I have some data on large language models specifically, which is kind of a subset, if you will. It's ETR data and said, to what extent do you anticipate your organization will use generative AI, LLMs, directly from the companies developing them versus embedding, embedded in existing vendor offerings. The number one response was, I'm not sure. So, by far. But for those who felt like they had a handle on it, it was very much balanced. It's kind of equally, some say all from the direct, others say all embedded, but most are saying it's going to be in balance. When you look at what they're doing with these large language models, you talk about monetization, they're largely monetizing through productivity improvements. They're saying, well, it's helped me write copy. It's given me drafts. It's saving me time. Better customer service, chatbots, etc. As opposed to I'm directly creating revenue from my data. Now, maybe that's how it really plays out. But I feel like there will be some genie out of the bottle that people go, aha moment, where people say, okay, I actually can monetize this more directly, create revenue versus just cutting cost and be more productive. Yeah, absolutely. And I was thinking about this the other day, that chat GPT is a great concept, right? And it's easy to see some of the simple use cases, helping people be more productive. But I feel like we've only seen the tip of the iceberg in terms of the capabilities of that technology. And I think as technologists, I think we yet don't know the full set of capabilities that that technology will be able to deliver for organizations. We're at the beginning of an era, I think, that this collision of all these multiple different worlds, but the data is really the key part where they're going to be able to monetize it. And I think the big piece of it is also that people are looking at it and the personas that you're talking to, the people who are actually buying the software and the hardware are changing. And I think that if you give them too many knobs, that also gives them a chance to misconfigure things and cyber resilience becomes an issue for them. Is that what you're also seeing? Absolutely. We're seeing the prevalence of the app developer, the data scientists become not necessarily the final decision maker, because at the end of the day, IT is still very involved in the decision-making process and carries a lot of the budget still, despite what some people might say in the industry. But you're absolutely right. These new personas, they don't want to deal with the plumbing underneath everything. They just want the easy button. They just want it to work. I've got a job to do. I got to write my code. I got to test it, push it through the pipeline and publish it to production. That's all they care about. It's interesting, Pete. It's all forever in this industry, we've had abstraction layers. I mean, virtualization was an abstraction layer. Kubernetes, we talk about super cloud is an abstraction layer. Now we bring AI to the equation. On the one hand, this is like the good and the bad. The abstraction layer makes things simpler, but then as you get into it, you can do more with it. You can push more data. You push it harder and harder and harder and all of a sudden there's all this complexity. Of course, it creates great opportunities for yet another abstraction layer. I wanted to ask you about a little bit about road map fusion and OpenShift. IBM's done an amazing job of keeping the culture of Red Hat going. You used to work there, right? But it's a fundamental part of IBM's strategy. It's go to market. It's capabilities. It's platform build out. So what does the future look like in terms of capabilities that you guys are developing in collaboration? Yeah, you're going to continue to see refinement around this concept of supporting not just containerized applications, but also virtualized apps. And there's a really good foundation there that Kubernetes has laid for us to be able to do that. So you're going to see that probably more in the short term. But you're also going to see this concept with Kubernetes. One of the bad things is the multiplication of clusters in a Kubernetes environment. And how do I manage that? And there's some stuff that we're working together with Red Hat on to enable the ability to manage hundreds of clusters. And we've got customers, one customer case in point, 650 Kubernetes OpenShift clusters that they're running. And that takes it to a whole other level in terms of how do I manage that many clusters? How do I do upgrades? How do I ensure that I've got fixes applied consistently in that entire environment? And so that's another thing. And what's really important about that, this whole multi-cluster concept, is providing those consistent services that I talked about, the Fusion 5, the persistence, resilience, security, mobility, and cataloging in a consistent way across all of those environments. This is where we start talking about things moving from just storage to data and data that you can consume. Yeah, so we're going to talk about that in our next segment. I'm actually excited to talk a little bit about how AI fits in that hybrid environment. Just give you another quick stat, talking about this equilibrium or this balance. When you ask people what percent of the customers are all in in the cloud, the number is like 14%. So it's not that large actually. They're all in. Of course, smaller companies, etc. This is maybe a little bit biased toward mid-sized to large companies. But if you ask them what's that going to look like in three years, they say the same number, 14%. So it's a hybrid world. Yeah. Pete Bray, thanks very much for coming to theCUBE. Thanks, guys, as always. All right, keep it right there. We'll be back to talk about AI in a hybrid cloud environment. You're watching the IBM Storage Summit live from theCUBE's Palo Alto Studios. We'll be right back.