Google is working on developing a new AI chip to improve the efficiency of Gemini.
Google's AI assistant, Gemini, runs on its own proprietary Gemini servers. It has now been revealed that Google is designing a new server chip to improve the operational efficiency of these Gemini servers.
Google Plans New 'Frozen' Chip to Run Its AI Models Much More Efficiently — The Information
Google is working on a new AI chip designed to make Gemini more efficient | TechCrunch
https://techcrunch.com/2026/07/20/google-is-working-on-a-new-ai-chip-designed-to-make-gemini-more-efficient/
According to information obtained by The Information from sources close to the company, Google is developing an AI chip called 'Frozen v2' internally, which is scheduled to be released in 2028. The report states that Frozen v2 could be 6 to 10 times more efficient than Google's existing AI chips when comparing the number of tokens generated per unit of power.
When technology media outlet TechCrunch contacted Google about the report, Google commented: 'Our teams are constantly researching and experimenting with new innovations to deliver the best performance and efficiency to our users and customers. Not all projects make it to market, but this thorough exploration is at the core of our full-stack approach. By co-designing hardware and software from the ground up, we ensure that systems are integrated and highly optimized for real-world workloads.'
AI companies are actively developing their own AI chips to improve the efficiency of their in-house AI models and address the global shortage of AI computing power. TechCrunch points out that as concerns about AI-related spending grow and the market fervor for AI cools down, the efficiency of AI chips is becoming a key selling point for technology companies.
In fact, OpenAI, which is competing fiercely with Google in the AI market, announced 'Jalapeño,' an inference chip optimized for large-scale language models (LLMs) developed in collaboration with Broadcom, at the end of June 2026.
OpenAI announces 'Jalapeño,' an LLM optimal inference chip jointly developed with Broadcom - GIGAZINE
In early July 2026, it was reported that AI company Anthropic was in discussions with Samsung to develop its own custom AI chip.
Reports indicate Anthropic is in discussions with Samsung to develop its own custom AI chip - GIGAZINE
Related Posts:
