Loading summary
IBM Representative
The thing about AI for business, it may not automatically fit the way your business works. At IBM we've seen this firsthand. But by embedding AI across hr, IT and procurement processes, we've reduced costs by millions, slash repetitive tasks, and freed thousands of hours for strategic work. Now we're helping companies get smarter by putting AI where it actually pays off, deep in the work that moves the business. Let's create smarter business. IBM
Nathan Hager
Bloomberg Audio Studios Podcasts Radio news
Karen Moscow
Shares of IBM trading near their lowest level since November 24th September of 2024. The company dialed back its full year sales forecast after a pretty steep drop in mainframe sales which weighed on results. Joining us from New York is the co host of Bloomberg's the Close, Romain Bostic alongside IBM CEO of In Krishna
Nathan Hager
remain Arvind, you've seen the reaction amongst investors here. Some concerns here about that lowered sales forecast overall as well as softness in software. You've characterized this as basically a shortfall for one quarter that is limited to capex sensitive areas of the portfolio. But that's still a meaningful area of your portfolio. Were sales coming into that quarter? Was that pipeline overstated?
Arvind Krishna
I don't believe so because when we look at all the deals that didn't close, I think we have done enough verification, including with the clients, to know that they were very real and it was a reprioritization of the CapEx spend. At the end of the quarter this was pretty confined to, I'll call it the Fortune 100, the deals that were within a subset of those. Now one third of what didn't happen had has already come back. So that tells us that this was a reprioritization and those deals were very real as opposed to us being optimistic in our projections and remain I would also add, I think that maintaining our free cash flow tells us that we have levers around productivity and conviction and confidence in the business and that is also I think going to serve our investors well. But the next few months will tell
Nathan Hager
us that well on that cash flow figure. Yes, and that certainly pleased a lot of analysts and investors out there. That billion dollar number of project free cash flow growth, you maintaining the dividend as well. But you're largely done that so far by cutting costs. So that raises the question that if we are anticipating slower growth on the revenue side, does that mean more cost cuts are in store?
Arvind Krishna
So one the bulk of our cash flow growth over the last four years has actually been on adjusted ebitda. So that tells you that this is mostly through revenue growth and our model has always been that we are the last dollar is more productive and more profitable than the first dollar. So we've been growing revenue 4 or 5% and we've been growing cash flow up in the 7, 8, 9%. So that's kind of our model and we intend to keep maintaining that. Now right now if we drop revenue by one point because you said 4 to 5 instead of 5 plus, we can absolutely make it through productivity. Cost cuts is an interesting question. Cost cuts doesn't always come down to to people reduction in headcount. Our headcount has been more or less flat over the last many years. I think there's a lot of third party spend where we are going to get a lot more efficient with that third party spend than we have been always.
Karen Moscow
Bloomberg Tech is live on Bloomberg Television and Bloomberg Radio and we're speaking with the IBM CEO Arvind Krishna. Good morning. You want to focus on accelerating revenue growth and accelerating profitability and just really simply I'd love to hear what you're asking the team to do differently now in response to all of the factors that you outlined.
Arvind Krishna
So it really so if I look at our software business, 80% of it is already an annuity consumption opex based business. 20% of it is a capex business. If we think that the capex headwinds are going to continue but 80% is already growing at about 8% we want to put a lot more focus so we are going to direct a lot of the team with forward deployed engineers with people who are focused on deploying the software at clients much more technical help and make that 80% grow even faster. Products like Red Hat, Confluent, Hashi all fit that model. On the capex side we have to make sure that while we can continue to do it don't depend upon outsized growth on that side to go there. Then on the supply chain can we leverage all of our capability and supply chain to make sure we have enough distributed infrastructure in storage, in Unix systems that people can fulfill all of the demand because we came out of the second quarter with half a billion dollars of backlog in that part of the portfolio. So those give you an idea of the kind of changes that we are making already, not just for the rest of the half.
Karen Moscow
You summarize the state of the world beautifully. You said that customers shifted spending towards servers, storage and memory in late June. And when I posted on social media you're coming on the show, the question from the audience really simple did that trend continue from June into July and for how long do you expect it to last?
Arvind Krishna
We have not seen it in July, but I would tell you, look, the semiconductor pricing memory is up what, three to four times over the last 18 months. Networking infrastructure is up 60 to 80%. Fiber is up, connectors are up. I do think that some of these trends on the underlying components are going to carry on for some more time. That works its way up into those things. So people are going to have those prices. Now the question is, is that going to maintain a reprioritization of the CapEx spend? If I take the signal that a third of our deals closed then that says no, people are going to get more careful about how to spend their capex. If that goes into September because a lot of capex does get committed at the end of a quarter, not always through, then it means that that carries on for some more time. And that is why we gave a guide of 4 to 5% depending upon if that carries on we'll be at the lower end of the range. But if some of it comes back, then we'll be at the high end of the range.
Nathan Hager
So. So Arvin, I know there's a big focus right now on, on sort of the, the clients that sort of did not necessarily materialize in the quarter. With regards to the existing clients that you have, and I'm primarily referring to your mainframe business. I mean can you share like sort of what percentage of those clients are actually increasing their spending? Is that going up still?
Arvind Krishna
Absolutely. So the current machine is called the Z17. The Z17 has been from the beginning, which was May of 2025 to today at 130% in terms of the capacity growth compared to the prior machine. That's one very strong signal and that's an aggregate. Then you can say, hey, is everybody in there? Or Only a few, 85% of the clients are increasing their capacity as opposed to the 15 who are not. That optimization goes on all the time. And, and I will tell you this is probably the best that we have seen in a long time of what is going on there. I'm going to go ahead.
Nathan Hager
Yeah, software is going to, I mean look, we know the hardware capacity is there, software's lag, and I understand why it's lagging in the moment. And I'm going to ask you a question and forgive me if it's a bit unfair, but I look back to the 1990s and whether there are some parallels to some of the mainframe issues that IBM went through back then under a predecessor two or Three times removed from you, John Akers. And this idea that at that time we were going through this paradigm shift in computing and how companies spent and allocated their money and there was a lot of talk by the CEO then that the issues were temporary, that they were economic and that they were rights themselves. We know in hindsight that it was much more structural. Why should we now look at this major shift going on with AI and the computation involved in that and think that this time it's is different?
Arvind Krishna
Well, we have to look at what the clients are doing. I always start there. In our opinion is an opinion. What clients do is what matters. So when I check with my clients, those who do credit card authorizations, are you going to maintain the mainframe? And they go to. Well, the resilience, the amount of capacity, the unit cost of it being five to 15 times cheaper is important. The big difference from that time, early 1990s to today remain is the cost issue. You could not argue that the mainframe was cheaper on a unit cost basis for the workloads that are moving off than at that time the mid range computers, I'll call it the whole Unix. It wasn't really client server, it was Unix. Taking the workloads off in the early 1990s, that issue you have to look at. If there is a cheaper alternate for that workload, I'll sort of look at you and say that means in five to 15 years it will move off. But that's not the case right now. For the workloads that are on, we can show the clients that it's five to 15 times cheaper to keep it on the mainframe than not. But that's not all workloads, that is workloads that need protection, that need resilience, that need the burst capacity that comes. And so for those kinds of workloads, the mainframe is the architecturally superior platform. That was not the case 30 years ago for the workloads that did go off.
Karen Moscow
Live on Bloomberg Television and on Bloomberg Radio. This is Bloomberg Tech speaking with IBM CEO Arvind Krishna. I've been listening to you a lot recently on long form podcasts, some clips on the social media about your view of the world, what it's like to be a CEO in the domains that IBM operates in. And it's interesting how quickly it comes back to the macroeconomic backdrop. A lot of people looked at the guidance for the balance of this year and would say, and they do say on this show, memory prices are not changing. They continue to push higher. They look at the Outlook for growth. Could you just explain the data points you rely on that give you the confidence on the new guidance that you've given and whether it is actually achievable or it will be difficult to meet?
Arvind Krishna
Yeah. So Ed, I'll come back to. For ourselves we have to look at our demand pipelines and our yields and how we get things going. But I think for your audience, let's look at it this way. Number one most important is what is GDP growth going to be? We think that that's between 2 and 3% for the year. If I look at the globe and it is going to be consistent even in the Middle east, even in Asia where there is a lot more energy and disruption, we actually see a lot of growth. Tech is going to be I think two to three to four points above that. So that puts tech in terms of what the market is somewhere in the 5, 6, 7%. Then that comes back to what parts of our portfolio can play against that demand and what parts cannot. And so I look at the parts of the software portfolio, that's why I talked about the 8% growth in the 80% of it that can play right into that. Then I look at our distributed infrastructure that can play right into that. I think consulting will be that 1 to 3%. It is not going to be in the double digit growers but we see the demand and we see the signings and we see the clients leaning in to say they want transformational work done. That's how we kind of know beginning with the macro and then coming down. I make a prediction for for you. I think technology spend is going to become a larger and larger part of every enterprise's budget. Used to be 3%. Since remain raised the 30 years ago it's probably up at 5, 6% on average. I will not be surprised if by 2035 it's 10% of everyone's budget.
Karen Moscow
Give you a quick micro case study. Bloomberg reported that Starbucks is replacing some IBM tools on the software side with in house. Talk to that.
Arvind Krishna
So Starbucks is about a little over $2 million a year client for IBM. The portion they're replacing is a product called Triga that does real estate lease management. The version of it that Starbucks has is almost 10 years old. I am not surprised that they're replacing it because I've actually been describing publicly software which is largely interaction based and is based on ease of use as opposed to anything else can be easily replaced by AI and agents. And that is what is going on there. However, if I see other parts of Starbucks and maybe the ability for our Hashi portfolio or security portfolio. How about if I phrase it this way, I would not be surprised if Starbucks is a larger client next year than it was last year.
Nathan Hager
With regards to how this world is evolving. Arvin, I am curious just internally about your plans to hire, particularly when it comes to the technology technological side. Have you been able to keep pace with some of the other companies out there also trying to do what you do and pay some of the salaries that you have to pay?
Arvind Krishna
Well, we hired three times as many college hires this year than we did last year. And I think that given others seem to be backing off college hiring, it gives us an incredible ability to bring in great talent and to then grow them inside our company and to offer them great careers. I think if I looked at it last, we had I think 20 million resumes in our applicant database. So that gives us a huge field to go look at. So I mean I don't worry about bringing in talent. I actually worry much more about can we give them a great career and a great pathway because not everybody is cut out to do work that is going to be demanded because I don't think there's a lack of employment, but the nature of the work. If you can't use air tools.
Nathan Hager
Yeah.
Arvind Krishna
If you can't use the productivity tools, then it's going to be really hard for you to be competitive with your peers.
Nathan Hager
We always, we already know sort of the potential impact of what I means for the economy and for your business. There are a lot of people looking around the corner, including yourself to Quantum. Is that a viable business on the horizon or is that just a moonshot that you're hoping actually sticks well, when
Arvind Krishna
things are two years away, I wouldn't call them a moonshot. I think Quantum is now in the engineering realm as opposed to the science realm for the next five years. And I believe that by 2029 we will deliver a machine that does 100 million computations, a large scale fault tolerant quantum computer in the next two years. I think that's an incredible opportunity. I'll quantify it. And this is not just our work. A lot of third parties. By the end of 2030s, we think it's about $1 trillion total market opportunity in terms of value that Quantum will create. That is why we also doubled down this morning and announced that we bought HRL from GM and Boeing. And that is going to help us bring people with a lot of talent around materials as spintronics, quantum sensing and other sensor technologies to add to our own effort, so we are even more well positioned to go win in this market.
Karen Moscow
Arvind Krishna, IBM CEO, of course, alongside Bloomberg's Romaine Bostick. Thank you both very much. Get the news you need in just 15 minutes.
Nathan Hager
Start your day with Bloomberg Daybreak, the podcast with a global view on the stories that matter. I'm Nathan Hager.
Karen Moscow
And I'm Karen Moscow. Join us each morning for curated stories on current events, politics, business and foreign
Nathan Hager
relations, plus one conversation on the day's biggest developments, all in just 15 minutes.
Karen Moscow
Subscribe to Bloomberg Daybreak for a precise, thoughtful take on the stories that matter.
Nathan Hager
Listen to Bloomberg Daybreak each morning on Apple, Spotify, or anywhere you listen to.
Date: July 23, 2026
Host: Bloomberg (Karen Moscow, Nathan Hager, Romain Bostic)
Guest: Arvind Krishna, CEO of IBM
This episode features an in-depth conversation with IBM CEO Arvind Krishna concerning the company's latest financial results, ongoing infrastructure deals, broader market dynamics, and IBM’s outlook on AI, mainframes, talent, and quantum computing. The discussion is timely, following IBM’s lower sales forecast and market challenges, with Krishna providing insight into IBM’s strategic responses and innovations in a changing tech landscape.
"It was a reprioritization of the CapEx spend. At the end of the quarter this was pretty confined to the Fortune 100...One third of what didn't happen has already come back."
— Arvind Krishna (01:22)
"The bulk of our cash flow growth over the last four years has actually been on adjusted EBITDA...We intend to keep maintaining that."
— Arvind Krishna (02:36)
"80% of [software] is already growing at about 8%. We want to put a lot more focus so we are going to direct a lot of the team with forward deployed engineers..."
— Arvind Krishna (03:55)
"Semiconductor pricing memory is up what, three to four times over the last 18 months. Networking infrastructure is up 60 to 80%. Fiber is up, connectors are up..."
— Arvind Krishna (05:31)
"For the workloads that are on, we can show the clients that it's five to 15 times cheaper to keep it on the mainframe than not...the mainframe is the architecturally superior platform."
— Arvind Krishna (08:19)
"Most important is what is GDP growth going to be? We think that's between 2 and 3%...Tech is going to be...two to three to four points above that."
— Arvind Krishna (10:32)
"Software which is largely interaction based and is based on ease of use as opposed to anything else can be easily replaced by AI and agents. And that is what is going on there."
— Arvind Krishna (12:21)
"I don't worry about bringing in talent. I actually worry much more about can we give them a great career and a great pathway because not everybody is cut out to do work that is going to be demanded..."
— Arvind Krishna (13:30)
"When things are two years away, I wouldn't call them a moonshot. I think quantum is now in the engineering realm as opposed to the science realm for the next five years."
— Arvind Krishna (14:44)
On Mainframe Cost Efficiency:
"For the workloads that are on, we can show the clients that it's five to 15 times cheaper to keep it on the mainframe than not."
— Arvind Krishna (08:19)
On Quantum Computing’s Promise:
"By the end of the 2030s, we think it's about $1 trillion total market opportunity in terms of value that Quantum will create."
— Arvind Krishna (14:44)
On Tech Spending’s Growing Share:
"I will not be surprised if by 2035 it's 10% of everyone's budget."
— Arvind Krishna (12:11)
Arvind Krishna’s interview offers clarity on IBM’s multi-faceted strategy for sustaining growth amidst short-term volatility, underlining persistent investment in AI, mainframes, and quantum computing. Krishna’s candid assessments about industry shifts—especially in light of CapEx headwinds, client transitions like Starbucks, and workforce transformation—demonstrate IBM’s ambitions and adaptability in a rapidly changing technology landscape.