Guardian investigation finds apparent discrepancy between what tech company has said about its AI capacity – and the number of advanced chips it has in operation
The chips are quite small and some can be held in the palm of a hand. They are fundamental to the development of artificial intelligence models – and the world’s biggest technology companies need vast numbers of them to keep ahead.
Microsoft is one of them. And, on paper, it seems to have a problem. A Guardian investigation has found an apparent discrepancy between what the company has said about its AI capacity – and the number of advanced AI chips it has in operation.
It is not a small shortfall either. Microsoft reportedly targeted having 1.8m AI chips installed in its datacentres around the globe by the end of 2024. Nearly two years on, in the middle of a $280bn (£208bn) expansion, the company has 2.2m AI chips installed, according to internal documents seen by the Guardian. This is less than half the number some experts had imagined.
Put simply, the global AI arms race requires a massive build-out of datacentres that run on extremely expensive chips. The apparent discrepancy over the chips suggests Microsoft’s newest datacentres may not be fully operational or, if they are, they do not have the chips they need.
This highlights something even more fundamental about charting the progress being made in the development of AI technologies. The chips that power AI are made by Nvidia, one of the two most valuable companies in the world. Its supply chain is one of the most tightly held secrets in the entire industry.
With almost no exceptions, Nvidia does not report how many of these chips it sells or to whom. Its clients, the world’s biggest tech companies, in turn do not reveal how many they have. Without this information, it is very hard for anyone to know whether AI is booming or not.
Microsoft: a power vacuum?
In the past two years, Microsoft says it has built AI infrastructure at breakneck speed. Its chief executive, Satya Nadella, said last year it would double its global datacentre footprint by mid-2027. Since 2022 it has ploughed roughly $280bn into the land, buildings and computational infrastructure to build AI. This includes more than $41bn in the past quarter.
But it is difficult to estimate how many datacentres Microsoft has built with this money.
It is possible to assess the progress that the company is making by looking at what it has announced publicly, with a particular eye on the power it needs. Datacentres need electricity, so one way of estimating how many datacentres are operational is to add up the energy Microsoft has at its disposal – its AI capacity.
Microsoft’s own claims, set out in annual reports and quarterly earnings, suggest it has added 5GW of datacentre capacity over the past two years as part of its AI build-out. It says it now has hundreds of datacentres on five continents.
Five gigawatts is a dizzying amount of energy – it is four times the size of the largest datacentre park in Europe. But Microsoft’s total capacity should be even greater than this; it has been building AI infrastructure since 2022. How much greater is an open question.
In an investor presentation from 2024, Microsoft reportedly claimed to have 5GW of datacentre capacity already installed. That would suggest it could now have a total of 10GW of capacity. It is unclear if all of these are AI datacentres – some could be for other cloud services. But Microsoft’s own statements indicate that the overwhelming focus of its capital expenditures in recent years has been to build AI infrastructure.
Ten gigawatts of AI datacentres would suggest Microsoft should have roughly 6.4m graphics processing units (GPUs). Shaolei Ren, a professor at the University of California, Riverside, said Microsoft’s sustainability reports, which contain figures for its electricity usage and are published separately from its financials, painted a different picture.
He said these reports suggest Microsoft’s AI capacity in 2024 was probably closer to 1.2GW. But even this lower figure would indicate Microsoft would need roughly 4m AI chips – if it added 5GW of AI datacentres in the past two years.
“According to their own metrics, Microsoft could be correct. But it isn’t clear what they mean when they say they have added datacentre capacity. They are giving insufficient context,” Ren said. “The sustainability reports are audited by a third party. They have more credibility than announcements.”
An analyst who specialises in Nvidia said they thought Microsoft would have more chips, given its public statements. “They’re low to me. They’re less than I expected Microsoft would have,” they said.
Microsoft insisted the Guardian’s calculations were based on incorrect information. It did not offer any insight as to which of the Guardian’s numbers were incorrect or why. What is clear is that Microsoft’s build-out of AI capacity appears to be going far more slowly than its annual reports may suggest.
Ren said: “It may be plausible to secure or announce 1GW of power capacity within a single quarter on paper. But bringing that capacity online and actually using it for computing within the same quarter would be far more difficult.”
Sources within Microsoft say the company’s total number of AI chips has “barely moved” over the past year.
Some of the apparent discrepancy may be explained by Microsoft’s tie-up with OpenAI. The exact terms of their commercial partnership are not public, but this unit may account for some of Microsoft’s datacentre deployments, which would not be in the documents the Guardian has seen.
‘You may have a bunch of chips … you can’t plug in’
There is another factor: some of Microsoft’s big projects appear to be far from operational.
Take Microsoft’s largest AI development in the US, a pair of datacentres in Wisconsin and Georgia called Fairwater. In April, Nadella, Microsoft’s chief executive, said the Fairwater project in Wisconsin “is going live”.
Satellite footage of the building from Epoch AI, however, appears to indicate only part of it is operational. In May, Microsoft admitted to a Wisconsin newspaper that Fairwater was not yet online.
This is very common, said Ren. Initially it was a multi-gigawatt, multibillion-dollar investment. Three years later, only 300MW has been built.
The internal document also indicates Microsoft has fewer of Nvidia’s newest model of chip, the Blackwell, than one might expect given Nvidia’s public announcements. Last March, Nvidia’s chief executive, Jensen Huang, said orders for Blackwells from Nvidia’s top four customers – widely thought to be Amazon, Oracle, Microsoft and Google – amounted to 3.6m.
There was no breakdown given for this figure, but Microsoft has historically been one of Nvidia’s largest customers. If this was still the case, that should put Microsoft’s total Blackwell holdings at somewhere close to 1m chips. In fact, it has less than half of this amount installed.
Where are the chips, if not in the datacentres?
Nvidia’s balance sheets appear to indicate that it has sold a great many chips; it posted a revenue of $215.9bn in February. Has Microsoft bought these but not installed them? How many, and are all of them in its possession?
Nadella appeared to gesture at this question on a podcast late last year called All Things AI, where he talked about Microsoft’s datacentre build-out. The biggest problem, he said, was electrical power and building datacentres close enough to where power was located.
“If you can’t do that, you may actually have a bunch of chips sitting in inventory that I can’t plug in. In fact, that is my problem today. It’s not a supply issue of chips. It’s actually the fact that I don’t have warm shells to plug into.”
A Microsoft spokesperson said: “Over several decades, Microsoft has built a global infrastructure to meet rapidly growing customer demand for cloud and AI services. Our datacentres combine custom silicon, AMD, Intel and Nvidia chips across multiple generations with the networking, storage and systems infrastructure required to operate at scale.
“Microsoft does not report on the volume of specific chips in its AI infrastructure. The estimates the Guardian has shared with us are inaccurate, drawing the wrong conclusions from incorrect assumptions.”
Nvidia did not respond to a request for comment.
How to calculate numbers of chips from a company’s ‘AI capacity’
The world’s biggest technology companies give figures for their AI capacity in terms of power: gigawatts. One gigawatt powers between 700,000 and 1m homes. Meta says its controversial Hyperion datacentre in Louisiana will have 5GW of capacity. The UK company DataVita is planning a 1GW datacentre in Lanarkshire.
Converting these figures into chips means calculating how many chips can be run with that amount of power. The Guardian used the following methodology, reviewing these calculations with Abdeltawab Hendawi, a professor at the University of Rhode Island, and Ren.
To get a very broad approximation of how many chips there are in a datacentre, you could divide the power usage of that datacentre by the power usage of an AI chip – for example, an H100. A single H100 uses 700W. If Microsoft has 10GW of capacity, dividing this by 700 watts suggests it should have 12m chips.
H100s make up the bulk of the chips described in the internal document. It also indicates that Microsoft has A100s, which use less power, and Blackwells, which use more.
But this approximation does not account for several factors. First, datacentres have cooling systems and other equipment, which also use electricity. Ren estimates that in a given AI datacentre, 80% of the electricity goes to computer chips. This is roughly in accordance with figures from the International Energy Agency, although the number depends on the efficiency of the datacentre. Eighty per cent of 10GW would suggest 8GW may actually be in use.
This is slightly lower than Microsoft’s own figures for its datacentre efficiency, which appear in a 2024 sustainability report and suggest that 89% of the electricity in its new datacentres powers the IT systems, with an 11% overhead.
Second, not all the chips in a datacentre are AI chips. Instead, AI chips are fitted on to server racks with other computer chips, such as memory chips, that help them run calculations. A server with eight H100 GPUs uses a maximum of about 10kW of power.
Dividing 8GW by 10kW gives 800,000 servers, or 6.4m chips.
This is a conservative estimate, as in practice companies such as Microsoft oversubscribe their power capacity to some extent – putting more chips in a datacentre than can be supported by their IT capacity, said Ren.
