According to an internal document seen by the media, Microsoft aims to accumulate 1.8 million artificial intelligence chips by the end of 2024.
Microsoft is trying to make generative AI faster, better, and cheaper, but this effort is largely dependent on Microsoft purchasing chips and graphics processing units (GPUs), mainly from Nvidia.
According to this document, Microsoft plans to triple the number of GPUs it has in 2024.
Microsoft is partnering with OpenAI and is at the cutting edge of the generative AI boom, and staying at the cutting edge in this field will be expensive.
According to two people familiar with the matter, Microsoft is expected to spend around $100 billion on GPUs and data centers from the current fiscal year to the 2027 fiscal year.
Microsoft spokesman Frank Shaw (Frank Shaw) declined to comment on the grounds that listed companies must abide by a standard quiet period before announcing results. Microsoft plans to announce results in the next few days.
Analysts at DA Davidson estimate that Microsoft spent 4.5 billion US dollars on Nvidia chips last year. A Microsoft executive said this figure is roughly in line with Microsoft's actual spending.
Microsoft is working internally to design its own artificial intelligence chips to reduce dependence on Nvidia, but some employees are skeptical because Microsoft is many years behind Nvidia and technology is advancing so fast.