Google is reportedly developing a new artificial intelligence chip designed to significantly improve the efficiency of its Gemini models, according to a recent report that has already moved the company's stock.
The chip, internally known as "Frozen v2," is being designed by Alphabet, Google's parent company, to serve as a more powerful and efficient processor for its in-house AI models. The Information first reported the development, citing anonymous sources familiar with the matter.
A Major Leap in AI Processing Efficiency
According to the report, Frozen v2 could deliver between six and 10 times greater efficiency compared to Google's current AI chips. This improvement would be measured by the number of tokens generated per unit of power consumed — a critical metric for AI workloads that require enormous computational resources.
The chip is reportedly slated for release sometime in 2028, placing it on a longer-term development timeline within Google's broader hardware roadmap.
Google did not directly confirm or deny the existence of the Frozen v2 project when approached for comment. In a statement provided to TechCrunch, the company emphasized its ongoing commitment to hardware and software innovation.
"Our teams are constantly researching and experimenting with new innovations to deliver maximum performance and efficiency for our users and customers," Google 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 Broader Push for Custom Silicon
Google's reported chip development reflects a wider trend across the AI industry. Major AI companies have increasingly invested in designing their own processors as a strategy to make their proprietary models run more efficiently and to mitigate global shortages in AI computing capacity.
Efficiency has emerged as a key competitive differentiator for technology firms, particularly as concerns about the scale of AI spending have tempered the market enthusiasm that once defined the sector. Companies are also working to reduce their reliance on Nvidia, the chipmaker that has long dominated the AI hardware market. Nvidia's stronghold has left major AI developers heavily dependent on its products, motivating firms to explore alternatives.
Google is not alone in this push. In June, OpenAI announced its first custom chip, an inference processor known as Jalapeño. More recently, reports surfaced that Anthropic is in discussions with Samsung about a potential chipmaking partnership, further underscoring the industry's shift toward in-house and diversified silicon solutions.
Investor Confidence and Massive Spending Pledges
The news of Frozen v2's potential efficiency gains appears to have reassured investors. Following the publication of The Information's report, Alphabet's stock rose approximately 3% during Monday morning trading, providing a lift ahead of the company's earnings report expected later this week.
The positive market reaction comes against a backdrop of investor concern over Alphabet's enormous planned capital expenditures. Earlier this year, Google disclosed plans to spend between $180 billion and $190 billion to advance its AI strategy. With such substantial financial commitments on the line, the company faces pressure to demonstrate that its investments will translate into tangible returns.
A chip capable of dramatically improving efficiency could help address those concerns by potentially reducing the cost and energy demands associated with running large-scale AI models. However, with a projected release date of 2028, it remains to be seen how the project will evolve and whether it will ultimately move into production.
For now, the report signals Google's determination to strengthen its position in the increasingly competitive AI landscape — both through software advancements and the custom hardware that powers them. As the race to build more efficient AI infrastructure intensifies, developments like Frozen v2 will be closely watched by investors, competitors, and industry observers alike. If you found this article insightful, consider sharing it with your network to keep others informed about the latest developments in AI hardware.
Source: TechCrunch AI
