Microsoft is focused on making generative AI faster, better, and cheaper with a significant reliance on purchasing GPUs primarily from Nvidia. According to a document reviewed by Business Insider, the company has set an internal target to accumulate 1.8 million AI chips by the end of 2024. To achieve this goal, Microsoft plans to triple the number of GPUs it has by 2024, with partnerships well underway in partnership with OpenAI.
The cost of staying at the forefront of generative AI is high, with Microsoft estimating around $100 billion in future spending on GPUs and data centers from the current fiscal year until 2027. Despite the significant investment, Microsoft declined to comment on these details, citing a quiet period public companies must follow before reporting earnings.
Other major tech companies are also stockpiling GPUs, contributing to Nvidia’s impressive valuation of over $2 trillion. Meta’s CEO announced plans to purchase hundreds of thousands of Nvidia H100 GPUs in 2024. Meanwhile, Nvidia continues to innovate with its latest GPU, the Blackwell, along with the powerful H100s already on the market.
While Microsoft aims to reduce its reliance on Nvidia by internally designing AI chips, some employees are skeptical due to the company lagging behind Nvidia in this area. With fast-paced advancements in AI technology staying competitive requires constant innovation and investment.
As Microsoft continues to push the boundaries of AI technology, employees and insiders with valuable insights are encouraged to share their perspectives with Business Insider. Those willing to contribute can contact Ashley Stewart via email or secure message for further discussion.