International Business Machines Corporation (NYSE:IBM) Credit Suisse 24th Annual Technology Virtual Conference December 1, 2020 2:30 PM ET
James Whitehurst – President
Conference Call Participants
Matthew Cabral – Credit Suisse
All right. I think we’re going to go ahead and get started. I am Matt Cabral. I cover IT hardware here at Credit Suisse. And we’re very pleased to have Jim Whitehurst here with us. Jim is President of IBM. And he’s responsible for the company’s cloud and cognitive software organization, as well as corporate strategy for the business. And prior to IBM, Jim was President and CEO of Red Hat. So, Jim, thank you very much for carving out a little bit of time to join us.
Matt, thanks for having me. I’m really thrilled to be here.
Q – Matthew Cabral
Perfect. Maybe just to kick off the conversation, I want to go back a little bit and sort of the tail end of when you were at Red Hat, you were considering selling the company. Can you talk a little bit about just what appealed to you about the combination with IBM and what you think that provided from a strategic level as you thought about different options for the company?
Look, I wouldn’t say we were considering selling, but what I will say is the challenge we were facing at Red Hat was we saw the need for a hybrid cloud platform. And when I say we saw that need, we were kind of duplicating what we did with Linux. So, back 15 years ago, with Linux, we saw a need to have a common platform that ran across every type of hardware. So, literally, same bit. So, whether that was a Dell server or an HP server, et cetera, et cetera, et cetera, there was value in having a common platform that every piece of hardware is certified to and every piece of software certified to. So, if you’re running SAP, every three years, you knew you could do a hardware refresh and you weren’t stuck with one vendor from the old Unix days.
So, we established that platform, and I’ll say Red Hat’s the default choice Linux. And as we all know, platform businesses add a lot of value because everybody buys it because everybody bought it last year, everybody got it last year, because everything certifies and so it builds on itself.
And so, we saw that same need in the cloud world, which is you’re going to have workloads and you have deployment options, which are now primarily public clouds, right? So, having a common infrastructure that runs across, that’s really, really valuable.
The problem is, how do you kind of create that network effect to get a platform going because if nobody’s using the platform, then no one’s going to certify to it. If people don’t certify to it, no one is going to use the platform. So, it’s that same network effect that you need to get started.
And so, we have this platform OpenShift, which is now the center of IBM’s hybrid cloud strategy. But at the time, it’s just Red Hat. And so, we’ve made good progress with Amazon and Microsoft, not a lot of tractions with ISVs. And then IBM came and started discussing with us. And all of a sudden, we think about trying to get that network effect going, having one of the largest ISVs commit to putting its software on the platform and having one of the largest and most credible enterprise sales organization get behind the platform, we saw as a way to get the platform going. When we win this platform and become the default choice hybrid cloud platform, that obviously is a very massive opportunity that has a lot of incremental value.
And so, the simple way I explained it to my board, it was less about additive synergies, and it was much more, hey, by ourselves, we have – I’m making up the numbers – 10% chance as Red Hat of landing this platform, and if we do, it’s worth a tremendous amount of money. As part of IBM, our percentage chances of winning go from, whatever, 10% to 60%. And so, the synergy is less additive and more of a substantial increase in the percentage likelihood that we can win that.
And frankly, the other people who were interested in Red Hat were larger in the hyperscale business. Let me just say that. And so, their model is less about let’s land this hybrid platform and it’s how do we help suck workloads from on-premise on to the cloud. And there’s nothing wrong with that. But IBM strategy of truly being hybrid to provide choice, I think, is unique relative to other hyperscalers. And so, the fit in terms of building and landing a new platform with IBM and where IBM wanted to go with strategy was – it’s a really tight fit.
And I want to come back to that notion of building out the hybrid platform in a minute. I guess – I think if memory serves me right, the deal close and I think it was July of last year, so we’ll round up, we’ll call it a year-and-a-half almost that you’ve been at IBM, just talk a little bit about what your role is now at the company and just what’s surprised you the most since you got onboard.
My role now, I run – we call it cloud and cognitive software. So, I run the software group and corporate strategy. And by the way, I don’t actually run Red Hat now. I’m the Chairman of Red Hat. But Red Hat directly reports to Arvind because there’s need for separation and we can certainly talk about that because of Red Hat’s relationship with IBM competitors. So, we didn’t want to have Red Hat report to someone who is responsible for IBM software portfolio. So, I’m responsible for that. So, it’s really about driving the future direction of the software portfolio.
And then, as corporate strategy, it is frankly helping both, no reimagine, but define the strategy and help execute across the integrated value of IBM against this hybrid cloud and AI opportunity, which is the core of the strategy.
So, started off in July laying out the strategy around hybrid cloud and AI, and now driving execution across the components of the portfolio that we’re keeping.
Got it. And I think IBM has been a proponent of hybrid first for a while. Certainly been accelerated since you came onboard. One of the things that’s happened this year since the onset of COVID is we’ve really seen a rapid acceleration in just adoption of public cloud. I think it’s a force that’s been underway for a while, but really kind of turbocharged this year. Can you just talk a little bit about what the role of public cloud is in sort of the steady state for enterprise IT longer term and if a year like this has changed your perspective at all?
Well, certainly, a year like this is accelerating the move to public cloud, right? As people are looking at net new workloads, they’re more likely to go on public clouds because people aren’t in their own data centers as much as they were, I don’t know if that changes the overall kind of percentage long term of workload that’s going to go into public cloud. I think what we’re seeing with kind of a common platform that kind of abstracts where workload runs, workloads will naturally fall where they are going to be most economic, and public clouds are going to be a large percentage of that. Is it 40% or 60%? I don’t know. But it’s probably going to be somewhere around half or slightly more could end up on public cloud.
And that’s not just because of economics of public cloud. That’s also because it’s where a lot of innovation is happening, whether that’s directly the cloud providers offering – or basically ISVs offering their services on their clouds or companies like Snowflake is a great example, or other companies that are kind of cloud first in how they’re offering their products. So, you’re going to see a lot of net new workloads drive that way. That will be somewhat offset by a lot of net new workloads that will leap out of the data center the other way into the factory floor, into the airline hub, into the retail establishment. And especially around AI and sensory where data has gravity, that data can’t get all the way back to a cloud economically for decisions to get made. So, you’ll see a big push out that way.
And then, you get the big transactional workloads that today are the traditional data center. The cloud really wasn’t designed to run well. And so, a lot of those will remain, and we think that’s the traditional data center. So, while, certainly, a large chunk of the growth will go towards public cloud, you’ll also see a substantial amount of growth kind of out into the business itself at the edge. And then, still a solid business because of the economics of a data center for a whole set of workloads.
Got it. So, coming from Red Hat, obviously, open source was critical to what it was that you guys were doing over there. Can you talk a little bit about just the importance of open source software to enterprise IT and just what that means now to IBM going forward?
Yeah. Because I really do think that we have the best of two worlds, with the expertise from Red Hat around open source, and then what we can uniquely do at IBM. And let me kind of unpack that a little bit. My shorthand way to talk about open source, [indiscernible] talk about it in marketing material, but it’s basically, if a problem is relevant to Web 2.0 companies, there is probably going to be a robust, pretty advanced set of open source projects around it.
So, if it’s problem that Facebook has, there’s going to be great technology around it. So, something like Hadoop, well, these big data problems were problems that Web 2.0 companies had, so they solved it. The big problem, some of those technologies are almost too big for enterprise use cases.
But anyway, there are these technologies out there around AI, around DevOps, around analytics, a whole bunch of areas. And really trying to compete against, frankly, a byproduct of Facebook, it’s out there for free is difficult and the model of user-driven innovation for those types of problems is hard to compete with. In other words, the people were solving a problem every single day on the line to add a feature to timeline, right, that’s a great way to solve that set of problems. And so, both the fact that it’s free and it’s being driven by people at the front – leading edge of problem solving, there’s incredible open source technology out there. And so, for a lot of core infrastructure, we’re going to see that technology out there.
Now the problem with open source is it was never meant or written for enterprises in mind and the people writing it really don’t care about the product lives and back porting and patches and stuff that a traditional enterprise would have.
So, that was the business at Red Hat built. So, we had this incredible capability to take open source projects and make those commercial, kind of supported, secure, reliable kind of projects.
But here’s the problem with that. And the reason I come back to that problem – is it a problem that a Web 2.0 company has, is that’s not every problem out there. So, let me just take AI, for instance, because it starts talk a little bit about IBM’s portfolio. Web 2.0 companies typically develop their application stack after we knew, ‘hey, let’s not have our data tied to the application, let’s have it nice and separated.’ Right? So, the idea of data governance and data cleansing not real relevant to a Web 2.0 company, but relevant to any traditional enterprise that’s more than 25 years old.
So, we will take and ingest the great open source AI/ML engines, the analytics tools in our products because we know very well how to take those things and life cycle the open source, but then augment with things like data governance and data cleansing. Or importantly, most of the AI that we hear talked about today is really ML. It’s machine learning. So, how you take a neural network and do learning? Phenomenal to identify pictures of cats, phenomenal at a lot of things enterprises need to do, you can’t make a loan based on that because it’s not auditable, and therefore, the regulator’s won’t let you do it.
And so, we’ve invested in technologies around AI that is auditable, so you can track bias, so you can demonstrate to regulators, so you can use it for those types of workloads, right?
So, we take the power of open source, augment it with things that enterprises need that open source doesn’t want to do and then we can kind of broadly offer that. So, open source is and will continue to play a large part of what we deliver, but it really is a combination of the lifecycle support and secure open source using our capabilities to do that, combined with augmenting that with technologies that aren’t coming out of open source to try to offer more holistic solutions for clients.
Let’s build on that a little bit. I know you’re not running Red Hat on a day to day. But I guess as IBM and Red Hat specifically is trying to build out, OpenShift is sort of that platform for hybrid, talk a little bit about sort of where kind of Kubernetes has brought containers as a technology to and just kind of that need for a more robust enterprise grade set of features and services on top of there and sort of that that maturity curve for OpenShift that you see going forward?
Yeah. Well, we’ve made tremendous progress. I think when Red Hat – I may be slightly off these numbers, not too far off. When we closed the deal, Red Hat had about 800 OpenShift customers. OpenShift is our kind of brand name for a collection of projects, including Kubernetes and Linux, and it’s basically the Kubernetes platform. We had about 800 customers. We’re well over 2,000 customers now. So, we’ve made tremendous progress with the platform.
And so, two pieces. I’ll talk a little bit about feature, then I’ll talk about kind of what IBM helps bring to the table. So, when I talk about lifecycle, for instance, open source projects basically never fixed bugs. It’s what we say. You fix it in the next release. You just release early release often. So, if you look at upstream Kubernetes where Red Hat’s the number two contributor behind Google, it’s updating all the time with bug fixes and issues.
The problem is, let’s say, you put an application portfolio on Kubernetes. And now, if it’s a platform or stuff that you’re going to update every week, it’s probably no big deal because you can keep up to date with the latest version of Kubernetes. But a lot of enterprises will build a set of applications. I’m going to run this for three to five years and I don’t necessarily want to have developers full time dedicated to continuously updating those types of things.
The problem is, Kubernetes itself is – other than Red Hat, no one else has the capability to lifecycle Kubernetes for a long time. Not to pick on anybody, but to use a big example, a great partner of ours is Amazon. But Amazon with their Kubernetes service will very clearly say in the documentation, we support N minus 2 releases. So, in other words, they will support two old releases because after that the new features drift, it’s hard to bug fix, right? Because it releases every six months, Amazon will support a Kubernetes application. And their documentation is pretty clear about – it’s likely your application will break beyond that if you’re not updating it, right? Well, we commit to a five year life. And over time, we extend those in the same way we’ve done with Linux. So, we have engineers on staff who can bug fix and support these platforms for a longer period of time. So, as an enterprise, you can plan. You know how long you’re going to have to support before you’re going to have to think about rolling forward.
And then, we work with enterprise clients saying, well, these applications, you’re probably not going to have problems, but these will. So, all that lifecycle work that an enterprise beads that we do on top of open source is what we do to be able to say, ‘hey, here’s a platform that you know is safe, secure, reliable, we obviously build in a whole set of security features, we lifecycle the whole piece together, including the Linux security updates.’ And people forget, the container has the – that’s getting in the weeds, but the userspace of the operating system in the container where literally 98% of security vulnerabilities are. So, if you’ve not been able to lifecycle that well, you have an issue.
We’re able to basically take open source and deliver enterprise great products out of it. Then IBM brings like a level of credibility that – I almost couldn’t imagine. You asked me early on, what are the things that surprised me. I never quite answer that question about IBM. Well, one of the things is just the credibility and depth of the relationships it has with its clients. Like one – just got a couple examples. Verizon is building their 5G, the mobile core for 5G on OpenShift. We always had a good relationship with Verizon at Red Hat. They always liked our technology. I don’t think they would have ever bet their 5G core on Red Hat. Well, they’re happy to do that with IBM because of IBM’s heritage of managing and supporting over decades kind of mission critical type systems. Or Schlumberger, another example, has a large application originally built on a public cloud, wanted to take it to clients – different customers around the world want to use different clouds. So, they’ve kind of backed up and are putting out doing it on OpenShift because it can run on premise or across into the other clouds. Again, I don’t think Schlumberger would have made that bet on Red Hat. They’re very happy to make that bet on IBM.
So, the capabilities of IBM, both the relationships and the ability to support these large implementations that kind of go across the globe really strengthens our hand in a way I didn’t fully realize how valuable that is and one of the reasons we’re seeing such good traction.
Let’s talk about that a little bit more because the numbers you guys been reporting in terms of the growth trajectory for Red Hat has picked up meaningfully relative to where you were as a standalone. And maybe expand on that that point and sort of both the growth in OpenShift that you’ve seen, but also bring RHEL into the discussion. So, just where you’ve seen sort of the opportunity to expand that RHEL base into IBM’s broader swath of customers around the world?
Yeah, well, we’re seeing both. Obviously – honestly, we probably had too much traction – or too much focus on the new thing with Kubernetes and OpenShift and probably not enough on Linux, but we’re seeing great traction across both of those. And again, I think a lot of that just comes down to the credibility of having IBM behind it. But importantly, it’s also the services capability to land workloads on the platform. Yeah, this is one of the chicken in the egg problems that we always had, is we always went to the big system integrators and said, hey, build a practice around this new thing. Back in the day, it was JBoss. Now it’s OpenShift container platform. And rightly so, the big GSIs would say, where’s the demand, show me the demand, and I’ll build a practice around it. We’d say, no, but you’ve got to build the practice now to generate the demand. And so, you kind of get this chicken and the egg that’s kind of natural when you’re trying to build a platform.
GBS, so IBM services is building a substantial practice and capabilities around Red Hat Linux and OpenShift and Ansible. So, it’s broader in the portfolio of Red Hat technologies. And that capability is driving a whole bunch of demand. And of course, that demand begets more demand. And so, it kind of builds on itself. So, having the IBM services capability as well to help build both capability as well as workloads on the platform is a big driver of that growth, and something that I think would have been harder to do at an arm’s length with a GSI.
One of the reasons we obviously – GBS, the consulting business is a key part of the portfolio is we’ve kind of rethought the portfolio going forward because there is such kind of tight synergy across – between the product business and the services business there.
And by the way, it’s also in the research business. So, one of the other things I’ve been blown away by is, as you can imagine, it’s harder to refactor an application to go from the stateful monolithic things to containerized micro services. So, we’ve had IBM Research building a whole set of tools that help do that.
And so, just kind of quickly, just to give you a sense of that, so if you think in our AI portfolio, you may have heard something called Project Debater. It’s kind of our follow on to winning Jeopardy, which is to have Watson basically be able to debate a championship debater. So, it can scan all this natural language stuff, understand meaning and context and do that.
I know that sounds like, okay, well, that’s a neat parlor trick. There are real world use cases beyond what I’m going to talk about in terms of natural language. However, computer code similar. So, the ability to be able to look at code and be able to say, okay, how do I containerize this has many of the same problems because you’ve got to look at this line and interpreting context. And that context might be a library that’s being called 50 lines above that. But because we built that capability to do natural language processing, we can apply that to computer source code, our ability to be able to scan a portfolio and say, ‘okay, of your 1,000 applications, these 850, we can do 98% of the work to containerize them automatically. We had a large win – I don’t know if the name is public, I’ll say in transportation, it’s a company that’s volumes are down because of the pandemic as you can imagine and it’s using this as an opportunity to do a substantial migration of applications while volumes are lower.
And really, the reason we wanted – and that’s a good example, it’s OpenShift running on a combination of Azure and AWS. So, unfortunately, it’s not IBM’s cloud. It’s two other clouds. But the reason – and GBS is doing the migration. GBS won the migration because of IBM Research and the work that we were able to do to be able to analyze that set of applications and be able to accelerate how fast the client can move, right? And then, they can move to OpenShift on either cloud. They won it for other reasons too. I don’t want to – we have good people too.
But the key point there is just our technology allows enterprises to move faster. And most enterprises, when they go into cloud – going to cloud means moving applications to cloud. And so, if we can help them do that, it puts us in a pole position to be kind of the strategic partner for clients as they look to move to cloud.
You touched on a little bit that sort of install base of legacy applications that are sitting in sort of that traditional on prem, they’re sort of running on top of the middleware they were built on however long ago, and it’s a pretty involved process to move those application. Maybe talk a little bit about what you see as sort of that migration process. You guys have something called Cloud Paks. Just what is that? And what is it that you guys are trying to accomplish with Cloud Paks as we start to think about that playing out going forward?
Cloud Paks are a broad concept. We’re taking all of our software and we are containerizing it and putting, we call it, Cloud Paks. So, they’re collections of what would have been before disparate products, containerized that can run anywhere.
So, number one, we are kind of eating our own dog food, saying we put this on OpenShift. And by putting on OpenShift, you can now run this anywhere. So, you want to run whatever the Cloud Paks for data and our analytic tools on Amazon have added. It runs there because it’s supported on OpenShift and OpenShift is supported there. Run it on premise, run it on Google. So, first off, we’re making our software much easier for our clients to run wherever they want to run it. By the way, that goes across chip architectures as well. We can come back to that because there’s some really exciting things around abstracting chip architectures.
But then as you said, so you get a lot of people, say, running WebSphere, right? And so, over time, they want to go to the new world, but you don’t necessarily want to cut over a day one. That can be a multi-year journey. And frankly, they’re probably pieces that never make sense to actually move for the performance characteristics of those.
So, how do you think about that problem? By creating these Cloud Paks, we can have a Cloud Pak that enables an enterprise to say, I’m going to buy one SKU and whether I want to run WebSphere in its current form or I want to run OpenShift and a set of cloud native tools, doesn’t matter. It’s one kind of product move that will – so we’re trying to eliminate this kind of bubble cost of moving from one to the other. But by enabling that to all the kind of – basically acquired together and you can choose when and how you decide to move across those, as well as providing really good tooling to help enterprises do that and manageability across that kind of set of applications.
We’re also continuing to drive feature velocity into that because you start observing more problems there. So, RPA, we did a recent acquisition there. We continue to drive the integration portfolio and open source integration that’s in OpenShift natively in other open source tools, there’s still a real need for incremental tooling that we have in the portfolio. So those things will continue to kind of coexist where and how the needs exists.
I think importantly for us, all of a sudden, by putting this on OpenShift, we’re not as limited to, I’d say, IBM platforms. Not that we were ever exclusively limited, obviously, when certain these other technologies run on other platforms, but as OpenShift is becoming kind of by far the leader of the hybrid cloud platform and we are the early adopter, putting whether it’s our Cloud Pak for security or integration or application or data on OpenShift, we are dramatically expanding the base on which we can sell into with the set of technologies, which is – so we can kind of get into specifics of what we’re doing in security, what we’re doing in data or what we’re doing an integration and application development. But importantly, we’re doing it on a platform that spans just kind of a lot larger footprint as we continue to go forward.
I’d love to dig into the specifics. Before we do, I guess, where are we in sort of that adoption curve of Cloud Paks across the portfolio? And thinking about it in terms of the broader cloud native platforms business, I guess, what is it going to take for it to sort of pull along more of that ex Red Hat piece sitting within that that organic cloud native platform for you guys?
Couple ways to answer that. First off, I would say we’re still in the early-ish innings. But what we’re seeing so far is very positive. And much of IBM – and we’re continuing to move towards more subscription type models and consumption. But we still have a lot of ELA and perpetuity models in there. We’ve done a really good job of getting Cloud Paks with those clients. And now we’re in the process of driving consumption against those, and we’re getting really good uptake.
So, I’d say we have a greater focus than we’ve had in the past on ensuring not only do we get the product sold to get it installed and used, and we’re seeing some really nice uptake in clients where we’ve kind of gotten it sold and up and going. Of course, these platforms, you need to kind of build the new applications for then the volumes to come. And so, there’s always a lag in there. But we feel really good if we look at kind of number of clients and kind of the initial workloads and the adoption and take off there.
To your broader question, maybe a so-so analogy to use. But VMware is an interesting analogy where VMware ESX is actually free. But by landing the hypervisor, it saw the management challenges associated with managing a virtualized data center. And so, you asked, well, why did VMware kind of disrupt the Big Four in management in the past, well, it’s because they went into the platform, they solved problems first.
We’re a little bit in that situation with the container platform. We are more likely to see the problems earlier on. So, whether it’s driving feature velocity and security or integration application development or in analytics or in management, by kind of having the pole position in the kind of the largest share today of Kubernetes workloads, it allows us I think to identify areas to continue to accelerate and invest in a differentiated way that are hard for others to do. We just announced, just an example, in Stata. As enterprises start – an acquisition and then Stana. So, as enterprises start having more applications containerized in micro services, well, imagine – well, suspend disbelief and say, in five years, you have an enterprise that has literally 4 million containers instantiating its application portfolio. Well, how are you doing application performance management, just basic metering, monitoring, et cetera. Those are areas that will continue to change as we learn more. And by being in the pole position at the platform, we’re most likely to see those first and be able to invest into those and accelerate our ability to offer the best management tools or the best AI and analytic capabilities on a containerized infrastructure.
So, there’s the initial, I’ll call it, first order of leveraging kind of the momentum we have in Kubernetes leadership today, those platforms, but also our ability to identify and accelerate into opportunities quickly because we’re the leader in the platform.
Let’s talk a little bit more about some of those kind of specific Cloud Pak offerings that you have. Let’s talk about security first. Just help us get a sense for what it is that you’re trying to accomplish with Cloud Paks for security and what that provides to the customer base relative to what they had historically.
First off, we are taking our existing offerings and containerizing them into Cloud Paks. So, number one, you can now run this where you want to run this, right? And so, that provides value and flexibility that just didn’t exist before for existing clients using these technologies in existing ways, right? It also allows us to with a more modern architecture, accelerate just our standard feature velocity that we’re driving into the platforms because we have a more modern architecture. And again, we can extend the kind of the footprint in which we can help our clients because we’re on kind of a broader platform that can run across clouds, et cetera, et cetera. And then, we can drive feature velocity very specifically associated with issues associated with new workloads. Right? So, whether that is container security and inspection, whether that’s new threat vectors that are coming in kind of related to these more modern platforms and, again, because we are there kind of early with the platform, we have an opportunity to observe and kind of drive into those areas.
And similar question around Cloud Paks for data. Maybe just help us understand sort of what that is and how that helps customers with those sort of on premise applications, what are those custom apps that they’re looking to bring to more of a containerized environment going forward?
Well, first off, again, you can take IBM software and run it wherever you want to run it. But now we’re bringing the analytic capabilities anywhere from database all the way through to AI, open scale which is our kind of anti-bias kind of set of technologies and where it’s making it much more easily exposed to the developer, right? So, the problem that you had with a lot of these, when they are kind of big monolithic pieces of software, you get a sales cycle and all the friction associated with that. Okay, you decided to buy it, let’s have a big implementation, et cetera, et cetera. Whatever, nine months later, you can start to see value out of it. By making the whole stack, A, more modular, but also kind of kind of more bundled together, it makes it much easier for developer to say, hey, I actually want to leverage open scale in this application in a much more frictionless way. So, first off, we’re making it easier for people innovating in the company to get exposed to our technologies because they’re just getting easier in that. We’re also making it easier to be able to run those things everywhere.
And then third, we’re able to drive a faster pace of feature velocity because we put it on the platform. We announced something called AI Ops where we’re saying, wow, you’re trying to run a modern cloud native environment, well, to do a really good job of automation, you’re going to have to use AI and learnings around that. Well, guess what, we happen to have those sets of technologies. So, our ability to get that product out and up and going quickly greatly enhanced because it’s just the overall architecture of having a containerized and a more modern environment.
So, again, it’s really – we can expose our functionality kind of more easily. We could extend to more platforms because OpenShift runs everywhere. And we could accelerate our own feature velocity on those. And that’s kind of generically true across kind of all the kind of the major Cloud Paks offerings.
Specific to those applications sitting on top of your middleware sitting on premise, just thinking through sort of the economics of what that shift looks like from a customer point of view. So, if they’re sitting with $1 of spend on let’s call it WebSphere on premise today, I guess the process and the path of bringing that on top of OpenShift, I guess what happens to that dollar? I guess, does it grow? Does it shrink? Does it stay the same? I guess what is that sort of mean to you as that sort of critical mass starts to move to more of that modernized environment going forward?
Well, frankly, generally, it grows because there’s a greater integration with the whole stack. So, the idea of kind of, well, what was once your – running on it. Generally, it had the Java Virtual Machine and then, obviously, the components that make up WebSphere. And oftentimes you had a different third party management suite that was actually providing visibility and maybe it was running on Windows or something else. So, there are a lot of other components in that stack. And what we’re able to do with OpenShift, is be able to offer a more holistic stack to address those workloads, which actually does kind of increase the addressable market. So, if you actually look, in absolute pricing, a Cloud Pak is a little more expensive, but you’re getting the entire infrastructure, not just the WebSphere component. So, the dollar goes up a little bit and not kind of greatly, 25%. Depends a little bit on the kind of the use case and the components around it. But in that go up, you are replacing a lot of components that you may have needed to buy separately from kind of other people in there. But most importantly, you want to win the share of net new workloads these are coming online and that’s the core that we’re working to do, which is, okay, you now – I got this set of WebSphere applications, I now want to run those in a more modern environment. So, I’m going to buy a Cloud Pak that has OpenShift. Now, well, I have OpenShift now as my container platform. My 50 other workloads, whether they’re going to use WebSphere or not or more likely to go on OpenShift because we’ve kind of seeded that as the core platform on which you’re putting things.
So, same thing. All the way down to you’re going to buy Sterling, our supply chain stuff, right? Now you have OpenShift in there, you’re more likely to kind of continue to put more applications on OpenShift. So, there’s a broad seeding capability that that brings us as well.
I want to broaden out the discussion a little bit. You mentioned a little bit earlier, you guys started sort of rethinking the portfolio and sort of what belongs within the wider IBM ultimately led to the decision to spin off the managed infrastructure services business. Maybe just help us understand a little bit more the thought process that went into that and why you think that’s the right path forward for the two businesses if ultimately that transaction is completed?
Yeah. Let me be clear, this didn’t start off with the – do we want to jettison the IS business or not. It started off with, let’s look at the portfolio and what do we think fits well in kind of this day and age.
And just a little bit of context to rewind, IBM for years has sold things and added things as we think is important. And the IS business, that business was created back in, call it, the late 90s and grew through the early 2000s. One of the big problem that enterprises had was, I want to implement an ERP system in the context of client server side. So, I had all of this complexity and heterogeneity at the hardware level and best of breed software. So, I had all this heterogeneity and complexity at the ISV level. And remember, Nicholas Carr wrote his famous article, IT doesn’t matter, it’s not a source of advantage.
So, when you put all that together, you had CIOs and CEOs who would say, hey, I just want to have a best in class ERP system. It’s not about competitive advantage. Who do I trust who can go take all this heterogeneity and complexity and bring it all together and give me a solution? Okay, that’s IBM. Here are the keys, go to do it for me. And so, that was an incredible business. And it’s still a really important business because there’s a lot of complexity in a traditional data center that still needs to be managed.
But if you look at the world of hybrid cloud, which is what we’re pivoting to, it’s all about how do I take homogenous infrastructure and accelerate my pace of innovation on top of that. Technology is now about how you create competitive advantage. Nobody wants to hand the keys to anybody. So, I’m going to outsource my IT, right? This is how you interact with customers. It’s how you build new businesses. So, there’s certainly – there are synergies still with the managed services business, but there are dyssynergies associated with it too because it’s a business – it’s around how do I manage complexity, how do I drive variants out, which is very, very different than how do I innovate, how do I frankly inject variants in that’s part of innovation. Both IS business and IBM, a little bit of conflict and potential logical partners. And so, that just no longer seemed like the synergies outweighed the dyssynergies of just a different operating model.
Now, that’s very different than – I talked a lot about the consulting business before and the importance of being able to invest ahead of the curve in hybrid cloud, which is why GBS fits well. Same thing on the hardware business. We haven’t had a chance to talk about it. But one of the things I’ve both been excited about and really encouraged by is IBM has unique capabilities in hardware. And the problem for a long time with that set of businesses is the friction cost between somebody working on a mainframe, to take the extreme, our work on x86 was pretty large. So, you had different tool chains, you had different management models, all of that.
Well, now with OpenShift, we can have a common infrastructure that runs across from mainframe to x86 to ARM, to ingest GPUs, we’ve announced a lot with Nvidia, et cetera. And where that starts to matter is, let’s say you have a cloudy AI developer working on an ML engine on Amazon, but the data they want to run it on is being generated by your transactional system on the mainframe. Well, rather now than having to pull all your data off the mainframe to go do that analytics, because OpenShift now runs on the mainframe, you just click a button run at night when the mainframe is not running transactions, do all the analytics on the mainframe and get the results back off. Because you’ve removed the friction, because you have the same operating model, because you have the same development tools and the development tool chains removing that friction from the platform now enables people to say, oh, I actually want to use this platform for this purpose. I don’t want to use it for that purpose. Now it works both ways. So you can say I got this mainframe application, I’m going to run it on OpenShift and, over time, I can pull pieces off that don’t make sense around the mainframe. But the flipside is, you can start to say, well, now my cost of running the mainframe isn’t so high because I need fewer skills from developers and all of that. So, I may want to move workloads back on. So, where it makes unique sense, I think will actually go to reflect that better. So, all that to say, the reason the hardware portfolio stays is we do have unique features that we think we can better reflect or expose, I guess is a better word for it, to developers and to enterprises by having that in the portfolio and exposed by the common hybrid cloud platform.
So, as we look through the hardware, the systems portfolio made sense, the GBS portfolio made sense, the software portfolio made sense, really the managed service business is the one that just didn’t quite kind of fit as tightly in terms of the synergy associated with our hybrid cloud direction.
On that point about GBS in particular. And you mentioned a little bit earlier when we were talking about sort of standing up OpenShift as a new platform and the benefits that that provides. Maybe talk about more broadly what GBS brings to IBM. And you talked about consulting a little bit. Maybe weave into there as well, just what applications management means longer term and what sort of keeping that within the portfolio does in this sort of push toward hybrid cloud from here?
Several things. Again, when enterprises are looking to transform, I just want to say, this is coming from the Red Hat person, nobody ever buys an infrastructure platform to look at an infrastructure platform, right? It’s the pipes and sewers where you want to actually run application. So, it’s either about building new applications or migrating existing applications. And what we found is our deep industry expertise combined – really understanding whether it’s an airline or a bank or a telco, combined with our AI capability in particular, so we call it cognitive enterprise, but being able to kind of imagine what are these business processes need to look like or where you want to kind of think about going with automation on a factory floor, we add that capability from our industry expertise, we understand the possibilities because of our AI/ML or our analytics capability and research broadly and we have the platform on which it can run.
So, the importance of GBS is, first off, to be able to help people migrate applications, but also imagine and build kind of transformative types of applications. And then honestly kind of help run those as well. There are a lot of enterprises essentially with this, I’ll call it, “do over” with this new – I won’t use paradigm shift because it’s too used, but this next generation technology. You’re saying, I don’t need to build skills to kind of manage this. I can have a public cloud kind of manage the underlying infrastructure. So, I need somebody to build and help me run my application. So, I don’t need to do that. I don’t have to worry about that. Let me find partners who can do that.
So, it’s a core part of us instantiating, delivering and partnering with enterprises long term on truly the hybrid cloud, which is the applications that are running on this fabric which might run across multiple clouds or it might run on the factory floor.
And one of the biggest questions I’ve gotten from investors since you guys announced the potential spin is just this target for mid-single digit growth for the core portfolio. Maybe just help us understand a little bit of sort of what gives you confidence that that’s the right trajectory for this business and just the biggest drivers relative to where historically you’ve been that you think will ultimately get you there post spin.
I’ll do the rough math, at least the world according to Jim. I’m not the CFO, but the math I’m looking at. So, Red Hat’s growing just under 20%. I think since we’ve closed, it’s grown 19%, but it’s been accelerating. Round out to call it 20%. And obviously, that’s becoming a bigger part of the software portfolio. So, you have kind of the Red Hat portfolio, call it, high teens to 20%.
GBS, assume post pandemic, it was – or pre pandemic, it was growing kind of 2% to 4%. So, let’s assume, pick your number, 2% to 4% for GBS. I think if anything with its affiliation with OpenShift, there’s more upside than downside from its traditional trajectory.
Obviously, the IS business will no longer be there. And you have the software business. And the software business, this year, frankly, we had a whole bunch of – we can get into that, some dynamics on the software business side. But in general, we feel really good about being able to have mid-single digits growth given the portfolio on top of kind of the leading hybrid cloud platform. And then, systems is kind of more flattish, but you do the math of all that together, mid-single digits doesn’t sound hard to get to.
Got it. Got it. And one of the things that – Arvind has talked a lot about this is just M&A as part of the strategy going forward. Maybe talk a little bit about it. I know you’ve announced a couple of kind of smaller deals over the last several weeks. But just bigger picture, What’s interesting from an M&A point of view and just where are the biggest areas of focus going forward?
This is kind of my point of view. So, I’ll be clear, it’s a little bit more of my portfolio, but I’ll speak a little bit to others. So, the M&A agenda now feels a little bit like it did at Red Hat over the last few years. And what I mean by that is, when you’re on the bleeding edge of landing a new platform, in our case, whether that was Linux in the day or the open source middleware stack or the OpenShift, they’re all kind of holes in that portfolio that you observe that you need to go fill. But they’re generally, I’ll call it, smaller acquisitions because it’s a new platform. You’re not going and buying a big established kind of business. So, in the Red Hat vernacular, it was stuff like, oh, we’ve got to handle storage. So we bought the company behind Ceph. It was called Inktank and Gluster. And oh well, we need automation in a stateless or in an agentless way. So we went and bought Ansible. But there were smaller tuck-ins to drive feature velocity in the core platform, right? So, if you think about where we are at IBM now, we are in that same leadership position with the platform, but on a larger scale. And so, things like in Stata, so we observed that there’s kind of need for better observability driven towards scale container platform, there’s going to be things around security, there’s going to be things – or other areas of manageability. So, we’ll be driving a set of acquisitions to fill in around that in that platform to maintain kind of feature leadership around the platform. We’ll be filling in that same kind of sets of holes on the consulting side.
So, we’ve made a couple of acquisition. We’ve announced at least one there recently. You’ll see us continue to build capability around, A, how to get applications built or moved to cloud. Now, that could be SAP going to Azure, right? So, that doesn’t have to be always net new applications. We’re going to continue to build our portfolio with the leading SaaS offerings which are going to get integrated in and more standardized.
So, a lot of times, those are things that you’ll buy versus build. And so, you’ll see us continue to do those incremental. What I don’t see, and famous last words, but I’ll say, at this point running strategy, I don’t see like a big gaping hole where there’s this big area we need to go fill, right? So, there’s no other like Red Hat thing we need to go do of size and scale. At this point, it feels like, continue to drive up organic investment profile, which we can more easily do post spin because we are more kind of growth – and then combine with tuck-ins to augmentation against the current strategy. Again, there’ll be things in AI and ML or in security and others will need to buy and the same on the services side. But again, I don’t see the need for kind of really big acquisitions from where we are right now.
Perfect. Unfortunately, we’re out of time. I could probably go for another half an hour or 45 minutes. But really enjoyed the discussion, Jim. Thank you very much for spending the time with us.
Yeah, thanks for having me. It’s good to catch up.