Google developing new AI chip to boost Gemini efficiency
▼ Summary
– Alphabet is designing a new server chip, codenamed “Frozen v2,” for its Gemini AI models, with a planned 2028 release.
– The chip is reported to be six to ten times more efficient than Google’s current AI chips, measured by tokens per unit of power.
– Google did not confirm or deny the report, stating it researches innovations for performance and efficiency.
– AI companies are developing custom chips to improve efficiency and reduce reliance on Nvidia, which dominates the AI chip market.
– News of the chip boosted Alphabet’s stock by about 3%, as investors seek reassurance on the company’s large AI expenditures.
Alphabet, the parent company of Google, is developing a fresh server chip aimed at making its proprietary Gemini models run with greater efficiency. Codenamed “Frozen v2” internally, the processor is expected to launch sometime in 2028, according to a report from The Information citing unnamed sources. The chip could deliver a six to tenfold improvement in efficiency over Google’s current AI hardware when measured by the number of tokens generated per unit of power.
Google did not directly confirm the report when asked by TechCrunch, but it also did not deny it. “Our teams are constantly researching and experimenting with new innovations to deliver maximum performance and efficiency for our users and customers,” a company spokesperson said. “While not every project moves into production, this rigorous exploration is central to our full stack approach. By co-designing our hardware and software from the ground up, we ensure our systems are integrated and highly optimized for real-world workloads.”
The push toward custom silicon is gaining momentum across the AI industry as companies seek to reduce operational costs and secure supply amid a global shortage of AI computing capacity. Efficiency has become a critical selling point as mounting concerns over AI spending temper the market enthusiasm that once surrounded the sector. At the same time, major players are actively working to reduce their reliance on Nvidia, the dominant force in AI chips whose hardware has become a bottleneck for many.
OpenAI announced its first custom chip, an inference processor called Jalapeño, in June. Earlier this month, reports emerged that Anthropic was in discussions with Samsung about a new chipmaking partnership.
Investors have long worried about Alphabet’s massive capital expenditures, which the company has earmarked between $180 billion and $190 billion to support its AI ambitions earlier this year. With such a staggering sum on the line, Google must demonstrate that its investments will generate returns. News of the more efficient Frozen v2 chip appears to have reassured the market, giving Google a boost ahead of its earnings report later this week. Following the report from The Information, Alphabet’s stock rose roughly 3% on Monday morning.
(Source: TechCrunch)
