Photo: IG Bank
Google works on Frozen v2 AI chip for Gemini: Report
The search giant's custom silicon could deliver 6-10x efficiency gains over current TPUs, with major implications for the compute economy powering crypto and AI infrastructure.
Google is developing a dedicated AI server chip known internally as Frozen v2 to optimize the performance of its Gemini artificial intelligence platform, The Information reported Monday, citing people familiar with the project. The chip could be introduced as soon as 2028.
Shares of Alphabet (GOOG) advanced 3.5% following the news.
The report said the processor will complement rather than replace Google’s Tensor Processing Units, with its architecture designed to reduce internal data transfers, improve inference efficiency and allow Gemini to answer prompts more quickly.
The tech giant joins a growing list of AI companies investing in custom silicon to enhance performance and lower infrastructure costs. According to the report, Anthropic is also exploring the development of proprietary AI hardware.