Meta spends $10 billion on Google Cloud to expand AI
Meta Platforms Inc. has reportedly struck a deal valued at a minimum of $10 billion with Alphabet Inc.’s Google for cloud computing services, as per sources close to the situation. This agreement is part of the social media giant’s extensive investment in artificial intelligence. Facebook and Instagram’s parent company is set to invest a minimum of $10 billion over the next six years for access to Google Cloud’s servers and storage. This move aims to rapidly enhance its computing capabilities in the competitive landscape of AI tool offerings. Sources familiar with the matter, who requested anonymity due to the confidential nature of the agreement, provided insights into this significant deal.
The agreement signifies the inaugural significant cloud computing partnership between Meta and Google, which holds the third position in the market, trailing behind Amazon.com Inc.’s AWS and Microsoft Corp.’s Azure. Earlier reports highlighted the details of the agreement. Google has confirmed the deal; however, no additional comments were provided. Meta has chosen not to provide a comment. Mark Zuckerberg, the Chief Executive Officer of Meta, has committed to investing hundreds of billions of dollars in artificial intelligence and the necessary infrastructure to support it. Despite having over two dozen data centers, the company is in the process of building additional facilities, including a massive 4 million-square-foot project in rural Louisiana. However, some of these centers will not be operational for several years. Zuckerberg aims to gather the highest amount of computing power for each AI researcher, with a sense of urgency in his approach.
In a notable development, Google Cloud has collaborated with Meta previously, but this marks the first instance of Google serving as a formal cloud infrastructure provider for the company. In 2023, Google Cloud revealed plans to offer versions of Meta’s open-source AI model, Llama, via its app developer platform known as Vertex AI. This move is part of Google Cloud’s broader strategy to establish itself as a versatile, “one-stop shop” for AI solutions. The distribution deal allowed businesses and developers to leverage Google Cloud for seamless access to and development of applications utilizing Meta’s AI models. In the past, Meta has conducted several smaller experiments utilizing Google Cloud technologies.
In a note released on Thursday, analysts Mandeep Singh and Robert Biggar stated that the multiyear deal “validates” Google Cloud’s lower token pricing when compared to other hyperscalers. In a recent analysis, it was noted that “Meta is likely to focus more of its efforts on improving the reasoning capabilities of its Llama model, given the traction of other frontier models in search, coding agents and real-time summaries and language translation.”
