Topline
CoreWeave signed a new deal with Meta to provide $14.2 billion worth of computing power to the Facebook owner, the latest in a slew of multi-billion dollar deals to support AI demands.
Key Facts
CoreWeave, an AI cloud computing firm, announced it would provide Meta with more than $14 billion worth of computing infrastructure in a deal that runs until 2031.
CoreWeave would give Meta access to the GB300, Nvidia’s latest computing systems, to speed up the time it takes to train new AI models, Bloomberg reported, as part of the deal.
Meta has spent billions on AI and data centers, saying in April that it plans to double its spending on AI, causing its capital expenditures to reach as much as $72 billion this year.
Meta has an option to expand the deal until 2032, according to a filing to the U.S. Securities and Exchange Commission.CoreWeave, which went public in March with a mouth-watering IPO, has seen its stock more than triple in value as demand for its infrastructure from major technology companies soars.
The announcement comes just days after CoreWeave added to its existing deal with OpenAI to train its most-advanced AI models by $6.5 billion, bringing its contract total with the ChatGPT-maker to $22.4 billion.
Key Background
CoreWeave is a cloud computing company operating AI data centers in the U.S. and Europe, offering access to Nvidia’s highly sought GPUs for training and running large language models. The company’s revenue comes from renting access to its compute capacity, often through long-term contracts with large enterprise customers. CoreWeave has signed deals with major technology companies including Microsoft, OpenAI and now Meta Platforms, as the companies race to meet booming computing capacity demands.
Tangent
The massive swell in the valuations of AI firms has sparked concerns among investors about whether the companies are overvalued and the AI boom is a bubble. About $1.5 trillion will be invested worldwide on AI this year by tech companies, and this could potentially rise to $2 trillion by 2026, according to Gartner, Inc., a business and technology insights company. The Wall Street Journal reported that while bigger tech players have increasingly invested in AI infrastructure, they have also been joined by many lesser companies, which are mostly debt-financed – increasing the chances of a burst similar to the dot-com bubble burst in the 1990s.
Further Reading
