
Summaries like this, in your inbox every morning.
Sign up free →Google announced its seventh-generation TPU chip called Ironwood in November, claiming it delivers 4x better performance than the previous generation for training and running large AI models. Meanwhile, Anthropic committed to purchasing up to 1 million TPUs from Google in October 2025 in a deal worth "tens of billions of dollars," and Meta Platforms signed a separate multibillion-dollar deal in February to rent Google's TPUs for its AI infrastructure.
Unlike Nvidia's chips sold on the open market, Google built these TPUs specifically for AI workloads and kept them mostly for internal use until recently. Now Google is selling them to external customers—a shift that matters because Anthropic and Meta explicitly chose TPUs over Nvidia chips for their "strong price-performance and efficiency," meaning Google's chips do more work per dollar spent.
For business leaders and investors: Nvidia currently controls 81% of the AI chip market, but analyst Gil Luria estimates Google could capture 20% if it continues this push, potentially becoming the second-largest AI chip provider and shrinking Nvidia's dominance significantly. For companies building AI systems, this creates real competition—you now have a credible alternative to Nvidia that major AI labs like Anthropic and Meta trust enough to bet billions on.
No comments yet. Be the first to share your thoughts!
Log in to join the discussion




Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack