Google Announces Eighth-Generation TPUs with Up to 3X Faster Training and 80% Better Cost Efficiency

According to Google, the company announced its eighth-generation TPU chips on April 24, delivering up to three times faster AI model training with 80% better cost performance per dollar compared to prior generations. The new TPU 8t and 8i variants mark the first time Google has split its chip lineup into specialized models for training and inference, enabling the company to run over one million TPUs in a single cluster. Analysts at Citizens project Google will generate approximately $3 billion in TPU-related infrastructure revenue in 2026, jumping to $25 billion in 2027, driven by strong adoption from AI labs and enterprise customers including Anthropic and Meta Platforms.
Disclaimer: The information on this page may come from third-party sources and is for reference only. It does not represent the views or opinions of Gate and does not constitute any financial, investment, or legal advice. Virtual asset trading involves high risk. Please do not rely solely on the information on this page when making decisions. For details, see the Disclaimer.
Comment
0/400
No comments