Svmuu News: Sources familiar with the matter reveal that Google is developing a new server chip, codenamed “Frozen v2.” The chip will directly integrate the architectural design of the Gemini AI model to improve the operational efficiency of AI models and alleviate current pressures on computing power supply. Google The company hopes this new chip will optimize the inference process for the Gemini model, enabling it to deliver AI services to users at a lower cost and with greater efficiency. Industry observers believe that if Frozen v2 is successfully launched, it will further strengthen Google’s self-reliance in the AI infrastructure sector and serve as a key asset in its competition with NVIDIA and AMD for dominance in the AI computing ecosystem. (The Information)