Alphabet shares climbed 3% on Monday after The Information reported the company is developing a new server chip, internally dubbed 'Frozen v2,' designed to run Gemini models more efficiently. The chip would permanently embed parts of Gemini's architecture directly into silicon, reducing calculations and data movement required to answer queries. Google engineers project it could serve between six and ten times more tokens per unit of power than the company's newest AI chips, called TPUs. The project targets a 2028 deployment and aims to ease a major internal compute shortage that has fueled tensions and reportedly forced Google Cloud to turn away outside business. The development reflects intensifying competition in custom AI chip design as tech companies seek to optimize infrastructure costs and performance for large language models.
Google Engineers Project 6-10x Efficiency Gains Over TPUs
The Frozen v2 chip would permanently embed parts of Gemini's architecture directly into the silicon, reducing the number of calculations and amount of data movement required to answer queries, according to The Information. Google engineers project it could serve between six and ten times more tokens per unit of power than the company's newest AI chips, called TPUs, or tensor processing units. Frozen would become a more specialized branch of Google's custom-chip portfolio rather than replace its general-purpose TPUs. The company is targeting 2028 for deployment.
Google Addresses Compute Shortage With SpaceX Deal
The project is aimed at easing a major internal compute shortage that has fueled tensions and reportedly forced Google Cloud to turn away outside business. Last month, Google agreed to pay SpaceX nearly $1 billion a month to help bridge the gap and meet its enterprise compute commitments.
Frozen v2 Chip Requires Consistent Gemini Architecture
The chip would work with future Gemini models only if Google sticks with the same underlying architecture, according to The Information. Google reportedly currently views Frozen v2 partly as a trial run and does not plan to produce it at the same scale as its TPUs. Alphabet did not immediately respond to a request for comment.
FAQ
What did Alphabet announce about the Frozen v2 chip on Monday?
Alphabet did not make an official announcement. The Information reported on Monday that Google is developing a new server chip internally dubbed 'Frozen v2,' designed to run Gemini models more efficiently. The chip would permanently embed parts of Gemini's architecture directly into silicon, and Google engineers project it could serve between six and ten times more tokens per unit of power than the company's newest TPUs.
When does Google target deployment of the Frozen v2 chip?
Google is targeting 2028 for deployment of the Frozen v2 chip, according to The Information. The company reportedly views Frozen v2 partly as a trial run and does not plan to produce it at the same scale as its general-purpose TPUs.
Why is Google developing the Frozen v2 chip?
The project is aimed at easing a major internal compute shortage that has fueled tensions and reportedly forced Google Cloud to turn away outside business. Last month, Google agreed to pay SpaceX nearly $1 billion a month to help bridge the gap and meet its enterprise compute commitments.
