Home
Meta Strikes Deal with Google for TPUs to Power In-House AI, Challenging NVIDIA's Dominance

A major rivalry is intensifying in the AI chip sector, centered on the goal of "moving beyond Nvidia." In a significant development, social media titan Meta has entered into a multi-year, multibillion-dollar agreement with Google to lease its custom-developed Tensor Processing Units (TPUs) for building next-generation AI models.
This strategic move directly confronts Nvidia's long-standing supremacy in the AI chip market. For years, Nvidia has been Meta's primary supplier for model training. In fact, Meta announced just days ago its intention to acquire millions of GPUs from both Nvidia and AMD. However, by leasing Google's TPUs, Meta aims not only to alleviate computing power constraints but also to evaluate non-GPU alternatives within its proprietary data centers. Reports suggest Meta is even considering direct purchases of TPUs starting next year.
Google's Strategy: Customer and Competitor
The rationale behind this agreement is nuanced. Google Cloud executives have targeted capturing roughly 10% of Nvidia's annual revenue—approximately $20 billion—through TPU sales. To reach this objective, Google is not only partnering with investment firms to offer TPU leases but is also pursuing major clients like OpenAI and Meta by emphasizing its differentiated technology.
Interestingly, due to the robust demand for cloud-based GPUs, Google itself remains one of Nvidia's most significant customers. The company must make substantial investments in Nvidia's latest chips to stay competitive in the cloud sector, even as it promotes its own TPUs to challenge Nvidia's market share.
Market Chain Reaction: Driving Chip Prices Lower
This competitive struggle in the AI chip market is proving beneficial for downstream developers. Industry reports indicate that, thanks to the availability of alternatives like TPUs, OpenAI successfully negotiated a 30% reduction in its procurement costs from Nvidia.
As industry leaders like Meta adopt a more diversified computing strategy, Nvidia's dominant market position is under unprecedented pressure. The "arms race" for computational infrastructure is evolving from a pure capacity contest into a comprehensive battle encompassing architecture and ecosystem.
Related article
DeepMind CEO Hassabis: I sleep six hours a day, usually feel energetic around 1 a.m.
Fortune recently featured an interview with Demis Hassabis, CEO of Google DeepMind, revealing his unconventional approach to rest and productivity. Hassabis disclosed that he sleeps very little, structuring his waking hours into two distinct work blo
OpenAI, Anthropic Vie for Market Share Despite Revenue Shortfalls
Despite recent reports suggesting OpenAI missed revenue targets, creating pressure on tech stocks this Tuesday, private AI lab investors remain resilient. Seasoned backers have confirmed they will not reduce investment despite negative media coverage
California AV Compliance: A New Era of Tickets, Geofences, and 1M Miles
Guident operates an AuveTech shuttle in South Florida, managing a four-mile route in West Palm Beach and a one-mile route in Boca Raton using its remote monitoring technology. | Credit: GuidentCalifornia is redefining the regulatory landscape for dri
Related Special Topic Recommendations
Comments (4)
0/500
Meta和Google联手搞TPU,这操作有点意思。NVIDIA的垄断地位看来要动摇了,毕竟谁也不想一直被卡脖子。不过这种巨头之间的博弈,最终受益的可能是我们开发者,毕竟竞争带来的是更低的成本和更多的选择。坐等后续发展,看看这出戏怎么唱。🤔
Interesting move! So Meta's basically renting Google's TPUs to build their own AI chips? That's a huge blow to Nvidia's monopoly. I wonder if this will actually lead to cheaper AI services for us users, or if it's just a power struggle between tech giants. 🤔 The 'beyond Nvidia' race is getting real!
Meta這步棋走得真妙!跟Google合作TPU,不僅能降低對NVIDIA的依賴,還能加速自家AI模型的開發。看來AI晶片市場的戰國時代要來了,期待看到更多競爭帶來的創新與價格調整。這對整個產業來說是好事吧?👍

A major rivalry is intensifying in the AI chip sector, centered on the goal of "moving beyond Nvidia." In a significant development, social media titan Meta has entered into a multi-year, multibillion-dollar agreement with Google to lease its custom-developed Tensor Processing Units (TPUs) for building next-generation AI models.
This strategic move directly confronts Nvidia's long-standing supremacy in the AI chip market. For years, Nvidia has been Meta's primary supplier for model training. In fact, Meta announced just days ago its intention to acquire millions of GPUs from both Nvidia and AMD. However, by leasing Google's TPUs, Meta aims not only to alleviate computing power constraints but also to evaluate non-GPU alternatives within its proprietary data centers. Reports suggest Meta is even considering direct purchases of TPUs starting next year.
Google's Strategy: Customer and Competitor
The rationale behind this agreement is nuanced. Google Cloud executives have targeted capturing roughly 10% of Nvidia's annual revenue—approximately $20 billion—through TPU sales. To reach this objective, Google is not only partnering with investment firms to offer TPU leases but is also pursuing major clients like OpenAI and Meta by emphasizing its differentiated technology.
Interestingly, due to the robust demand for cloud-based GPUs, Google itself remains one of Nvidia's most significant customers. The company must make substantial investments in Nvidia's latest chips to stay competitive in the cloud sector, even as it promotes its own TPUs to challenge Nvidia's market share.
Market Chain Reaction: Driving Chip Prices Lower
This competitive struggle in the AI chip market is proving beneficial for downstream developers. Industry reports indicate that, thanks to the availability of alternatives like TPUs, OpenAI successfully negotiated a 30% reduction in its procurement costs from Nvidia.
As industry leaders like Meta adopt a more diversified computing strategy, Nvidia's dominant market position is under unprecedented pressure. The "arms race" for computational infrastructure is evolving from a pure capacity contest into a comprehensive battle encompassing architecture and ecosystem.
DeepMind CEO Hassabis: I sleep six hours a day, usually feel energetic around 1 a.m.
Fortune recently featured an interview with Demis Hassabis, CEO of Google DeepMind, revealing his unconventional approach to rest and productivity. Hassabis disclosed that he sleeps very little, structuring his waking hours into two distinct work blo
OpenAI, Anthropic Vie for Market Share Despite Revenue Shortfalls
Despite recent reports suggesting OpenAI missed revenue targets, creating pressure on tech stocks this Tuesday, private AI lab investors remain resilient. Seasoned backers have confirmed they will not reduce investment despite negative media coverage
Meta和Google联手搞TPU,这操作有点意思。NVIDIA的垄断地位看来要动摇了,毕竟谁也不想一直被卡脖子。不过这种巨头之间的博弈,最终受益的可能是我们开发者,毕竟竞争带来的是更低的成本和更多的选择。坐等后续发展,看看这出戏怎么唱。🤔
Interesting move! So Meta's basically renting Google's TPUs to build their own AI chips? That's a huge blow to Nvidia's monopoly. I wonder if this will actually lead to cheaper AI services for us users, or if it's just a power struggle between tech giants. 🤔 The 'beyond Nvidia' race is getting real!
Meta這步棋走得真妙!跟Google合作TPU,不僅能降低對NVIDIA的依賴,還能加速自家AI模型的開發。看來AI晶片市場的戰國時代要來了,期待看到更多競爭帶來的創新與價格調整。這對整個產業來說是好事吧?👍











