Google is working on a new AI chip designed to make Gemini more efficient

2026-07-20 · Source: AI News & Artificial Intelligence | TechCrunch · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, AI Hardware · Depth: Novice, quick

Summary

Alphabet, Google's parent company, is developing a new server chip, internally named "Frozen v2," to enhance the operational efficiency of its in-house Gemini AI models. Slated for release in 2028, this chip is reportedly 6 to 10 times more efficient than Google's current AI chips, measured by tokens generated per unit of power. This initiative aligns with a broader industry trend where AI companies like OpenAI and Anthropic are designing custom chips to reduce dependence on Nvidia, address global computing shortages, and improve cost efficiency amidst investor concerns about massive AI expenditures. Google plans to spend between \$180 billion and \$190 billion on its AI strategy. News of Frozen v2's potential efficiency gains positively impacted Alphabet's stock, which climbed 3% following the report.

Key takeaway

For investors evaluating AI company valuations, Google's "Frozen v2" chip initiative signals a critical shift towards in-house hardware development. This move aims to significantly improve Gemini model efficiency by 2028, potentially reducing operational costs and mitigating reliance on third-party chipmakers like Nvidia. You should consider how custom silicon strategies impact long-term profitability and competitive advantage when assessing AI sector investments.

Key insights

AI companies are developing custom chips to boost model efficiency and reduce reliance on external hardware providers.

Principles

• Custom hardware optimizes AI model performance.
• Chip independence mitigates supply chain risks.
• Efficiency gains address investor spending concerns.

In practice

• Evaluate custom chip development for large-scale AI.
• Prioritize power efficiency in AI infrastructure planning.
• Diversify AI hardware suppliers beyond single vendors.

Topics

• AI Chips
• Google Gemini
• Custom Silicon
• Hardware Efficiency
• AI Infrastructure
• Investor Relations

Best for: CTO, VP of Engineering/Data, Executive, AI Hardware Engineer, Investor, Director of AI/ML

Related on AIssential

• Google reportedly develops Frozen v2 chip for Gemini AI — Google is developing a highly efficient custom AI chip, Frozen v2, reflecting a broader industry shift towards proprietary silicon for AI.
• Google Plans New ‘Frozen’ Chip to Run Its AI Models Much More Efficiently — Google's "Frozen v2" chip directly integrates the Gemini AI model blueprint for 6-10x greater efficiency.
• Google's "Frozen v2" chip reportedly bakes Gemini's architecture directly into silicon for efficiency gains — Google's "Frozen v2" chip hardcodes Gemini's architecture for 6-10x AI inference efficiency, deploying 2028.
• Alphabet Stock Jumps After Information Report of New Google Chip — Google is developing "Frozen v2," a new AI chip 6-10x more efficient, integrating Gemini's blueprint.
• Google Signs $920 Million Monthly AI Compute Deal with SpaceX — Google's massive compute deal with SpaceX underscores intense, sustained AI infrastructure demand and GPU scarcity.
• Meta’s new AI chips will begin production in September — Major tech companies are developing custom AI chips to reduce costs and optimize for specific workloads.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI News & Artificial Intelligence | TechCrunch.