Meta, the parent company of Facebook, is reportedly in advanced discussions with Alphabet’s Google to secure a multi-billion dollar agreement for the use of Google’s tensor processing units (TPUs) in its data centers. According to a report by The Information, this collaboration could see Meta renting TPUs from Google Cloud as early as next year, with full integration into its data centers slated for 2027. Google has positioned its TPUs as a cost-effective and secure alternative to Nvidia’s chips, which have faced supply constraints. The report also suggests that Google aims to capture 10% of Nvidia’s revenue through its TPU business. Neither Meta, Google, nor Nvidia have commented on the matter, and Reuters has yet to independently verify the report. Meta has been a significant customer of Nvidia since 2022, utilizing its graphics processing units (GPUs) to train AI models and support its vast user base of over 3 billion daily app users. Earlier this year, Meta announced a $600 billion investment in U.S. infrastructure and jobs over the next three years, including the development of AI data centers.
