International Business Machines Corporation (IBM) Earnings Call Transcript & Summary

Hello, and welcome to the Fourth Annual Wells Fargo TMT Summit. My name is Phil Winslow. I'm the software analyst here at the bank. Very excited to have IBM joining us for the second day of this year's conference. And really going to delve down into Red Hat, and joining me is Paul. If we can bring Paul up on the screen. Paul, thank you for the time today, and looking forward to a really exciting discussion.

My pleasure. It should be fun.

Great. Now Paul, you've obviously had a rich career at Red Hat over 20 years now, culminating in your appointment as Red Hat CEO just earlier this year. Can you share us a bit about the journey that you've been on and part of driving Red Hat as a leader in the enterprise-grade open source market?

Sure. It's been quite a journey so far. I've been here, as he said, about 20 years now. I think I was employee 120 or so. So when I started at the company, it was purely consumer focused. We were selling box products in the bookstores. And we knew that if we really were going to make it, we had to move to the enterprise. And so that's -- that was the big pivot that we made as a public company, actually. So that was -- and we decided new product, new everything with it, and we moved that to the enterprise. And we really rebuilded the company. But even as we started with RHEL in the enterprise, it was really a commodity flex. It was really -- in fact, we used to even talk about RHEL in the beginning as 75% the function of UNIX. I guess we don't have the best marketing people back then. But that was the play in really getting people from Windows as well as from the very expensive hardware Unix combination to the x86 world. And -- but what happened over that 15 or 20 years was that because Linux was so available and so open, it became so powerful. And I would say probably 5 or 6 or 7 years in, all the innovation started happening around Linux just because of all those reasons. It was there, it was powerful, it was open, it could be sort of worked and massaged to work in new technologies. That's where cloud came from. All these things came out of the Linux world. And so Linux quickly moved from a commodity play to the innovation engine and more of a value play. And now you see these more complex solutions, whether it's cloud itself or whether it's things like containers and Kubernetes and all the things that go around it those are really looked at from a value perspective versus a commodity perspective. So that's really the big change that's changed over the years. I mean in terms of hybrid, because we grew up in the Linux space, we started working very early with our customers as they started to look at Linux-based clouds. And we recognize that the movement to cloud was not necessarily a flip the switch and you're going. It's all about the application. It's new applications, it's refactoring applications, it's re-architecting applications. So we knew the practical way was going to be hybrid because there's always going to be a set of applications that are going to stay on-premise and bare metal. We're going to stay as virtual machines on-premise. Customers are starting to build out their own private clouds with things like OpenStack, especially in the telco world. And now even as customers started to go to clouds, I think Microsoft had a lot to do with this because Azure grew so rapidly. They said, "Look, I might want to go to multiple clouds." And so everything for the last 8 years or so that we did around new products, extending current products, M&A that we did. My tenure at Red Hat, I think, has done 31 acquisitions. Everything was around the last 8 years around building out that hybrid cloud platform around Linux. And so for example, we bought 2 storage companies. We didn't buy storage to compete with NetApp and EMC. We bought storage so that the hybrid platform could have a storage layer to be able to access data no matter where the application ran. So that's sort of condensation of the 20-year journey to where we are now from a product perspective. And it's in its early stages, it's by no way even mid-stage yet.

Great. Now as a follow-up on that, let's drill in on cloud in our conversion. Can you talk about the dynamics in overall enterprise cloud adoption that you witnessed here in 2020? And particularly, how has COVID impacted enterprise's path to the cloud? And how do you expect this is going to change over the next few years? For example, what percentage of workloads are in the cloud today versus what this will look like?

Sure. I mean if you look at it, even Amazon, AWS started, what, 12, 13 years ago, something like that. And even back then, the message was every application is going to the cloud tomorrow. And there were even big companies out there. A couple of the banks that say we're going all in. I'm going to have all my applications in the cloud in 5 years. And not many if or any are even close to that now, not because cloud doesn't bring a lot of value, it's because, as I talked about earlier, the practicality of that. As people started to get into moving their applications to cloud, they recognized that, as I said earlier, because they were working alongside of us, that these are never going to go. These aren't ready to go. I'm not ready for these to go. I might want to go to multiple clouds. And so even over that time, the end users, the customers started to shift because to say it's going to be a hybrid world and -- because just for practicality reasons. Now I think that COVID probably accelerated that a bit a lot as well from what we're seeing with our customers because one of the problems customers have now, the hybrid world -- the cloud world and the hybrid world brings a ton of value, but it's complex. And the products we build, obviously, are to help make that simpler to move, manage, operate and develop, but it's still complex. So customers really are -- they don't have a lot of -- enough of the skill sets to manage all of these. And so that's where people like us come in right now. COVID has accelerated that, I think, because of the fact that, as I said, it's -- a lot of it's around skill sets and just being able to build and refactor these applications, it takes the right set of skill sets. I mean that's really -- also from our perspective, it's where IBM's worked with us and helped us a lot from their GBS side because one of the things on the GBS side that they're putting a lot of focus on is re-architecting, recoding and moving applications on the OpenShift platform as well. So that was good timing on both of our parts here. We've always partnered with all of the SI vendors out there. We still do, we still do. GBS, obviously, is one of the best, but that combination together, having the products, the platform as well as the consulting in order to do it, is what customers need help with right now.

Yes, obviously. Let's focus on Kubernetes. And we totally agree with you in terms of your sentiment of the accelerated adoption of cloud post-COVID-19. In fact, actually a 1 liner since March has been that no cloud strategies are officially a no go. So let's focus in on our conversation on containers, on Kubernetes as really the enablers of the industry's multi-cloud feature that you've been talking about. Now I think we'd all agree that dockers basically become sort of the de facto standard at the container layer. So let's focus on orchestration and scheduling for which obviously, Kubernetes is becoming rather synonymous. Can you just level set things for everyone on today's webcast describing what role do you see Kubernetes playing and why the industry is so abuzz about Kubernetes?

Well, I really -- I mean Kubernetes' path is exactly the same as the Linux path in terms of upstream, to community, to product, to downstream. Let me -- I'll explain more of them. Kubernetes has turned into really a substrate where services connect to. So compute, the Linux operating system connects containers, connects to Kubernetes. So I mean really, if you look at containers, not everybody really realizes this is that it's Linux. It's just Linux carved up in a different way. The host of the container is the kernels. And the user space of the operating system is part of the container. And so that's really what it is, it's Linux. The reason why Kubernetes could really even be developed is because Linux was open. Kubernetes APIs touch into the Linux kernel, they touch into the Linux user space, they're very much intertwined with each other. They couldn't have done that had they been working with a proprietary operating system where they couldn't munge it around in order to do that. But Kubernetes turns out to be the really the substrate that hangs the compute services, networking services, the storage services. It's really pretty light weight, actually. I mean -- and even the platforms out there today like OpenShift, it's much more than Kubernetes. Kubernetes is just one piece of it. And it really follows the same way as Linux, as I said. I mean it's all developed upstream. And if you looked at it, I'll use Linux as an analogy because everyone is more familiar with that. Same analogy, it's exactly the same. We all -- there's pre-distros out there. We all look -- come from the same Linux upstream kernel. We're all on the way to make products decide what file systems we're going to package, what developer tools we're going to package, what life cycle we're going to package, what bugs we're going to fix and what time frame. And so at the end of the day, that becomes a product. But because we all make these decisions, that's why RHEL is different from SUSE, which is different from Ubuntu. It's the same thing with Kubernetes. Everyone's distribution of Kubernetes comes from the same place, but they're all integrated differently. So EKS is not necessarily compatible with AKS, which is different from the Kubernetes that's part of OpenShift. Same core, same technology comes from the same place but different life cycles. And so Kubernetes plays that big role now to tie all those services together in the platform. But what's really happening in the hybrid world, platform stretched now. Bare metal, running on bare metal, running on VMs public -- private clouds, multiple public clouds, that's the new data center. So you don't want to have 5 islands of different implementations of things for your developers to develop to, your operators to operate and your security people secure. And that's where really, we come in and where hybrid comes in to work across that with these technologies that are all compatible with each other. You don't want to have to write your application for 1 cloud provider and never be able to run that application ever any in place else. You want to be able to maybe write it in one place and but be able, a year from now, run it in a different place. So that's what's really happening now as you're seeing that data center now becoming not necessarily -- not even necessarily, not at all, within your 4 walls. It's really spread out across these different footprints.

Sure. I mean to your point, unifying enterprise data centers with public cloud has obviously becoming the necessity, at least in our opinion. And we believe in the power of Kubernetes to ensure, we call it via parity, like a better term made between the 2 infrastructures, it's going to represent a lucrative business opportunity for software vendors. However, one of the pushbacks I often get from investors is that, hey, since Kubernetes is open source, then why don't companies just DIY Kubernetes as opposed to paying a subscription for a package distribution such as OpenShift. What's your response to that?

Some people do. I mean I think I just saw a paper, I think it was Salesforce. Did you see that paper? They did a DIY paper on Kubernetes. If you read through that paper, or I think it was a blog post actually. So if you read through that and look at what it takes to do that, there's probably 10 companies -- I mean Google does it, Facebook does it. But there's probably less than 10 or 15 companies on the planet that has that size, breadth of expertise to able to do that because you have to be very active in the upstream because that's where the development happens. And if you want to ever get anything fixed, you have to be active up there. You have to be a Linux, a deep kernel Linux developer. I think you have to be in and around all parts of Linux as a developer. So yes, you can do it. But just like I said, there's probably less than 10 or 15 companies on the planet we're servicing. We're helping the other 99.9% of the companies do that.

Yes. I just want to say, yes, normal humans can't DIY this. That sucks.

I mean one of the biggest problems in the industry in this space, the skills, it's finding the right skills. And so that's -- it's great to be a software engineer especially in this place right now.

Exactly. Exactly No, obviously, I agree with your response where DIY is pretty unlikely except for a handful of folks. But so let's focus on the competitive dynamics in terms of unpackaged Kubernetes landscape. The past 12, 24 months, we've seen VMware by Heptio and then Pivotal and then in Project Pacific and Tanzu Mission Control. Google released EGK and then Anthos and Red Hat, obviously, has a very long focus on Kubernetes as key to OpenShift. And it was bought by IBM and so on and so on. I can keep going. It's -- how do you compare IBM Red Hat strategy with those other vendors' Kubernetes strategies, i.e., what's most differentiated about OpenShift versus Tanzu versus Anthos?

Well, we'll get to that, but let's just start for just 2 seconds on the IBM side. So even when IBM first bought us, they had their own container platform, which means their own Linux platform and their own Kubernetes distribution. Cloud Paks was in IBM before we were at IBM, as a matter of fact, right? So what they decided to do early on in our tenure was they were going to retire that. And that was going to be all OpenShift. And so that's one thing. So all the software from the IBM Software group that runs on a container platform, it's exclusively on OpenShift. I mean it sounds like a really easy decision, and it sounds like a really easy thing to do, but it took a long time. It took a while to get there. We're there now. So that's one thing. In terms of the others, frankly, our biggest partners are our biggest competitors as well. I mean if you look at VMware, we've had a long-standing partnership with them. We run our virtual machines on their hypervisor for a long time, right? But I think what's happened is VMs are, although they're going to be around for a long time, forever, because it's a really long tail, I mean that's also part of that tail that I talked about. The virtual machine technology is now even moving into the Kubernetes world as well as CNV. But if you look at it, I think one of the things VMware sort of did is they never really focused on the application. In the virtual machine world, they were very good at the hypervisor, and they were very good at management of applications that were running as virtual machines. But they left the OS piece to us in Microsoft. That's where the applications run. On OS that's 2 things: interfaces to the hardware, interfaces to the app. So now that as we've moved to the container world, they've recognized it's going to be a container world, that's all about the app because it's about really giving you enough of the OS with the app to be able to traverse these different hosts. And so they bought -- I don't know if you said it, 6 or 7 companies. They went down the Pivotal pivot for a while. I think it's a lot of work to bring those together. Not to mention, they have to become a commercial Linux provider overnight. And we've sort of have a 15-plus year head start on that because it is RHEL in the containers. It's the same kernel, and it's much of the same user space. Every time we fix a security or any kind of bug in RHEL, it's the same fix in core OS -- RHEL core OS because it's the same code. So we know how to do this. We've got a lot of experience. In terms of the cloud providers, I think many of them sort of got caught off guard because now they're realizing on-premise is not going to go away. When you saw the Amazon announcement yesterday, right. They're coming on-premise in a big way as well as multi-cloud is something that's key in the whole hybrid world. So yes, now they're trying to come on-premise, the cloud providers. And so yes, we're absolutely watching that. But still, it's going to be part of that vertical stack of that. EKS is going to come on-premise, but that's not going to get you out of the EKS world. Then getting to the other clouds, I think, is going to be not just Amazon, all the cloud providers, working deeply on the other clouds is going to be a difficult proposition for them. Because we're neutral, we have deep engineering relationships with all of them, and that's what it takes. I mean OpenShift right now is it's in the Amazon console. It's the only non-Amazon service in the Amazon console is OpenShift. And so I doubt that AKS and Azure -- from Azure is going to get to that point or vice versa anytime soon. So I think that's the challenge they have is expand on-premise and expand vertically to the other clouds as well. So -- and it's still just a container world for them. Even then bringing EKS to on premise, there's still going to be a big part of applications that are run as virtual machines, and they're even run bare metal. So I just think we have a head start on that, but we never rest on our laurels. So it's -- we need to keep ahead of that.

All right. Let's switch gears then, sort of double-click on a few comments -- a few sentences ago. But it's on the hyperscalers and the competition on the hyperscalers. As you mentioned, AWS, Azure, Google, et cetera, all have Kubernetes services in their public clouds. How do you think about these what would you call managed cloud services versus deploying a Kubernetes distribution such as OpenShift? What are the positives and negatives? And why do you believe that OpenShift will resonate more with customers versus a Kubernetes service from a hyperscaler?

Well, those are 2 -- 3 different things, but I mean-- managed services is big. We have 2 big pushes beyond hybrid for the next 5 years for us. And one is managed services and the other is moving out to the edge, and we can talk about that later if you want to. But managed services, for some set of applications, is really big. We have a managed service on IBM cloud, managed service on Amazon and managed service on Amazon -- on Azure. As I said, we talked earlier about skill sets. The skill sets to actually deploy your own platform, manage our own platform. Although we have many tools for that, it takes skills. A lot of companies just can't find those skill sets. So they don't want to have to worry about the platform. And they just want to use the platform, and that's where we come in with managed services. So we've been in the managed service business across these for about 4 or 5 years. We've got our own operations center on this. We can tie all those together as one offering, so we can give you a managed service across all the clouds because we're on all the clouds right now. So it's really starting to be a big demand for that, especially from the skill sets perspective. And even everything else. Middleware, we've got the JBoss middleware portfolio. That's moving and move -- starting to move to completely as a service on top of the OpenShift platform. So -- and that will fit -- that fits with the managed service platform as well. Customers, developers don't want to worry about that. They just want to code for it. For a long time, more so -- I mean even now for a long time, we -- we've looked at OpenShift as our cloud platform. It just wasn't stationary. It just didn't run in one place. It could move. So we're actually -- what you're seeing with OpenShift, especially with OpenShift 4, is it's becoming more like the cloud platform, more hands off, more managed. In the past with Linux because the history of the Linux people was -- were much more deeply technical. So real Linux people wanted to fiddle with the platform. They want to make their own -- they want to set their own variables, make their own adjustments, all of that. Once we went to OpenShift 4, we got more opinionated on what the configuration should look like, how you should manage it, in many cases, did it for yourself. We even can call home with the platform now to get metrics. We can tell you we can tell you when you're going to have a platform problem before you know you have a platform problem. That's kind of what the cloud providers do as well. They just take all of that guesswork out, if that's what you want. So we think we have to do both sides of that. So I think that's really a big part of what managed services are all about everything as a service, and this is our platform to do it. And you have to be able to orchestrate from the platform up to do it, and that's what OpenShift is about.

Yes. And to your point before on EKS, and you and Azure Kubernetes services, that's sort of everybody's own flavor. Like if that form of Kubernetes and, okay, it might be able to run on their cloud, but what happens if you want to move it from Azure to Google? Well, it might be that same call it open source core, but just sort of like Java back in the day. Okay, yesterday, everything was a J2EE compliant application, but you couldn't move it from [ West Street ] to where [indiscernible] is, right?

It's true. It's not just the movement either. As I said with the Linux side, we all package -- it's a different environment, different set of development tools, different development environment, different operations tools. So now if you have 5 islands, if your developers are going to develop over here today and over there tomorrow, they have to switch development environments. If you're operations people, how they operate, it's going to be a different set of tools. Well, the biggest things in hybrid in our most -- one of our most successful product lines is Ansible for Automation. Because hybrid, it's a lot of value, but it's a lot of complexity. You have to automate across that. It makes it much easier to automate when you have a common platform where everything lives. It makes it much easier and predictable to do the automation. So I think that's an important part of it as well. I mean we've moved -- Linux turned into a box product. And it was -- if I look at our Linux days, what we were doing back then was so simple compared to what we're doing right now. It's just much more powerful, but it's also much more complex.

I'd say definitely harder, but complexity has gone up. I guess my next question is given the role that you just talked about playing with cloud providers and enterprises, how have you and the Red Hat team been helped by the acquisition by IBM? You had also been able to find sort of a balance in terms of maintaining neutrality in the market.

I'll talk about neutrality first. I mean going into the discussions on this, I guess, almost 3 years ago, 2.5 or so with Arvind and Ginni and Jim. From day 1, it was about neutrality because IBM's bet here -- IBM is a cloud provider, but their bet was an even bigger market was going to be the hybrid market, which was the path we took a long time ago, right? And so they recognized from the beginning, if you're going to be -- being a hybrid player with only 1 partner is not really a hybrid play. And so it was different for them, right? It was different for them. And so that neutrality in keeping that neutrality with the other partners is essential. I mean when we announced this, one of the first calls I got was from Intel because they compete on the hardware side. And we see Intel's road maps, we have a 5-year view of their road maps. It's really important that they guard that secret. We have to do that because we are the operating system. We had the show, and we have proven to them that we could keep that wallet and keep that on the Red Hat side, and we've got their comfort on that. One of our strongest partners is Microsoft, who obviously, Azure competes with IBM Cloud. We could never -- we work with the IBM cloud guys just like we do with the Azure guys, but we had never -- could never tell the IBM cloud guys what we're working on with the Microsoft Cloud guys and vice versa. Everybody accepts that on both the Red Hat side and the IBM side. So just having that understanding from the beginning is really, really important around the neutrality part of it. So I think that's the first. Now what was the second piece of your question. You said from neutrality to...

On how you benefit?

How we benefit. From the product side, we've accelerated. So for example, we were a small company. The decision I had to make on what we're going to do next for new products were very much driven by what the market would accept from us in terms of profitability and all of that, right? We get to IBM. They had a project going -- that managed multi-cluster Kubernetes. And it was all open source. We decided to -- with them to move that whole group about 200 engineers to the Red Hat side. So I mean it wasn't just, okay, stay in your IBM office, and we'll just put a Red Hat name on it. They actually moved offices, they moved job classifications different set of benefits, et cetera, to the Red Hat side because we wanted it to be integrated with the rest of Red Hat. That probably -- it was on our road map, it probably accelerated us by 4-plus years in order to do that. So huge benefit on that. And it's a real key part of our hybrid strategy from the go-to-market side. We always had a good tradition and reputation of selling great to the technical world, but because we were a small company, we were just starting to get a part in the organization. IBM sells at the C-suite and has for a long time. To bring those 2 things together now where you can sell to the technical guys and the C-suite guys at the same time, bringing those together, it not only accelerates deals, but it can actually close deals. Instead of doing a small proof of concept and someone over here that can sign off on a small proof of concept, you look at a bigger maybe company-wide project because you have more support from it further up in the management chain. So that's helped us not only grow the size of our deals, but it's also helped us in reach that we just wouldn't have gotten to yet. When we were small and we were just competing with SUSE and Ubuntu, sort of grow at your own pace. The good news is everybody is on Linux and open source, the bad news is everybody, from a competitive perspective, everybody is on Linux and open source. So that's -- this was a needed growth engine for us in order to really recognize the potential we have.

Yes. Great. Well, our time is going fast. We've only got a couple of minutes left here. So I'm going to ask you a high-level question. So we're going to fast forward 5 years. Let's say, Paul, you and I are back up on stage, hopefully in person in Las Vegas.

I love craps.

As you said, we'll go play craps the night before that. But what do you think you're going to look back on and say, in 5 years, hey, look, this happened faster than I expected back in 2020 or this technology or trend was more transformative than I thought 5 years ago?

I think the biggest thing coming is out to the edge. I mean this is happening faster than it ever would. Now that people are saying, okay, I'm going to be running in these different footprints across different clouds, in my own data center, I wanted to look at 1 client. I also want to get my compute and my data closer. I want to get my data closer to where the compute is, whether that's the factory floor, the cell tower, whatever that is. But I don't want to leave that out there as an island. That needs to be part of my hybrid world. And so that's really what I think is in 5 years or less even. We're doing it today in some industries, especially telco. That's going to be the biggest thing out there is to be able to do that. And we even see OpenShift as a big part of that, but also RHEL is a big part of that. You might not need a container platform out there. You might just need a Linux platform out there. And so we're looking at a lot of solutions with our RHEL and our Linux experience with that as well. It's just got a connecting. It's going to be managed as one, it's going to be able to be automated under 1 umbrella, it's going to be updated installed under 1 umbrella. That's where a lot of the future growth is going to come is out there at the edge in all these different verticals out there.

I'm glad to hear you say that back in -- I think it was 2017. In January 2017, we wrote about edge. We didn't know what to call it back that then. We think this thing out here, and we were calling it microfog and it had a couple of other terms, but like microservices based, fog-based [indiscernible]. But it was like, I think it's going to be big, and then it was things like intelligent edge out of Microsoft that may go big and from there. But one of the things we talked about is sort of a extension of that sort of that cloud computing model across multiple tiers, sort of the deep edge microprocessor, basically refrigerator, et cetera, to point factory floor, even the edge of the Internet or the cell tower. So a continuum that spans multiple tiers, sort of the right workload in the right place relative to whatever that application is for.

Exactly. Exactly. So...

That's complex though. There's complexity in managing that. Yes, so...

It is. Automation is key. Automation is key.

Yes. Awesome. Well, that was a fabulous conversation. I could have gone on for another half hour or 2 or more.

Me too. It was great.

Appreciate the time, and you're invited back in 5 years, so we can push [indiscernible].

We'll see in Las Vegas. We'll see in Las Vegas.

Awesome. Thanks.

All right. Great. Buh-bye.