Google Develops Frozen v2 AI Inference Chip with 10x Energy Efficiency Improvement

Google is developing the Frozen v2 AI inference chip, which will incorporate parts of the Gemini architecture into the chip to reduce computation and data transfer when running models. The chip is expected to process 6 to 10 times more tokens per watt than the latest TPU, with plans for deployment as early as 2028. This move aims to alleviate the shortage of computing power, as the gap has forced Google Cloud to turn down some external customer orders. Frozen v2 will coexist with TPU, which can run different models, while Frozen v2 trades flexibility for efficiency. This is also a bet on architecture; if Gemini undergoes significant changes to its underlying design in the future, these chips may no longer be usable.

Disclaimer: This content is provided for general branding and informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online events, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets or to use any services. Crypto assets are highly volatile and may result in loss. WEEX services and online events may not be available in all regions and are subject to applicable laws, regulations, and eligibility requirements. You are responsible for ensuring that your use of WEEX services complies with local laws and for carefully assessing the risks before participating in any crypto-related activities.