Meta has signed a $14.2 billion agreement with CoreWeave for the supply of AI infrastructure. The contract runs for six years and can be extended by one year until the end of 2032. The deal is intended to ensure Meta has long-term access to the computing power needed for the development of artificial intelligence.
As emphasized by SiliconAngle previously, this represents a major step forward for CoreWeave. The company, which was relatively unknown just a few years ago, has quickly become a key player in the AI sector. CoreWeave operates dozens of data centers in North America and Europe and provides companies with access to powerful Nvidia GPUs that are essential for training large language models. According to the latest figures, CoreWeave has 28 operational facilities and is working on ten new locations that are due to be completed this year.
More extensive collaboration with OpenAI
The dependence on a few large customers makes the agreement with Meta particularly important. Until now, it was assumed that a significant portion of the revenue came from Microsoft. In addition, CoreWeave expanded an existing collaboration with OpenAI last week, bringing the total contract value with that party to $22.4 billion. AI startups such as Cohere and Mistral also use the infrastructure, although these are smaller deals.
CoreWeave CEO Michael Intrator stated that this agreement gives Meta access to the latest Nvidia GB300 GPU systems. These chips are essential for applications developed by Meta. Think, for example, of the integration of AI into new consumer products such as Ray-Ban smart glasses.
Meta has been investing heavily in data center infrastructure for years. In April, the company reported that capital investments this year will total $72 billion, most of which will go to AI-related projects. At the same time, it is investing in talent, spending large sums to attract leading AI engineers.
The latter does not always go smoothly, as Techzine reported in August. At that time, Meta Platforms froze recruitment within its AI division. This happened after a period of several months in which a total of more than fifty researchers and engineers were recruited. These were talents from OpenAI, Google, Apple, xAI, and Anthropic. This measure was accompanied by a reorganization of the AI department.