Alphabet Developing Frozen v2 Chip for Gemini AI Models, Planned for 2028 Release

Alphabet is designing a new server chip internally dubbed Frozen v2 to run its Gemini artificial intelligence models more efficiently. The chip is planned for release in 2028 and could deliver between six and ten times more tokens per unit of power than Google’s current AI chips. The Information reported the project, citing anonymous sources.

Google did not confirm or deny the effort when asked for comment by TechCrunch. “Our teams are constantly researching and experimenting with new innovations to deliver maximum performance and efficiency for our users and customers,” Google said. “While not every project moves into production, this rigorous exploration is central to our full stack approach.

Other AI companies have moved to build custom chips. OpenAI announced its Jalapeño inference processor in June. Anthropic has discussed a chipmaking partnership with Samsung. Google earlier this year said it plans to spend between $180 billion and $190 billion on AI infrastructure.

News of the Frozen v2 chip lifted Alphabet shares 3 percent in early trading ahead of the company’s earnings report later this week.