Brief IA

Google bets on Frozen v2 to revolutionize Gemini's efficiency

🤖 Models & LLM·Tom Levy·

Google bets on Frozen v2 to revolutionize Gemini's efficiency

Google bets on Frozen v2 to revolutionize Gemini's efficiency
Key Takeaways
1Alphabet is developing a new AI chip, Frozen v2, to enhance the efficiency of its Gemini models.
2The chip, expected in 2028, could surpass Google's current chips by six to ten times in energy efficiency.
3Tech companies are looking to reduce their reliance on Nvidia by creating their own AI chips.
💡Why it mattersThis initiative could transform Google's competitiveness in the AI sector, influencing its costs and market position.
Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisis ton rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

📄
Full Analysis

Alphabet and Its New AI Chip for Gemini

Alphabet, the parent company of Google, is developing a new server chip aimed at optimizing the performance of its artificial intelligence models, Gemini. This initiative seeks to make the operations of AI models more efficient, a crucial challenge in a rapidly expanding sector.

The Innovation of the Frozen v2 Chip

Internally dubbed "Frozen v2," this chip is expected to debut in 2028, according to a report from The Information based on anonymous sources. This component promises to be six to ten times more powerful than Google's current AI chips, particularly in terms of the number of tokens generated per unit of energy consumed. This advancement could represent a significant turning point for the processing capabilities of the company's AI models.

Reactions from Google and Strategic Challenges

When questioned by TechCrunch, Google neither confirmed nor denied this information. However, the company emphasized its ongoing commitment to research and experimentation to provide optimal performance for its users. Google stresses the importance of co-designing its hardware and software to ensure maximum integration and optimization of systems for real-world workloads.

The Trend Toward AI Chip Autonomy

In the current landscape, AI companies are increasingly looking to design their own chips. This trend addresses both the need to improve the efficiency of internal models and the desire to mitigate global shortages of AI computing capacity. Energy efficiency has become a crucial selling point, especially as AI spending raises concerns in the market, tempering initial enthusiasm.

Competition Against Nvidia

Historically, Nvidia has dominated the AI chip market, making major industry players dependent on its hardware. However, companies like OpenAI and Anthropic are now exploring alternatives. OpenAI recently unveiled its first custom chip, Jalapeño, while Anthropic is considering a partnership with Samsung for chip manufacturing.

Financial Implications for Alphabet

Investors have expressed concerns about the colossal expenses anticipated by Alphabet to support its AI strategy. Earlier this year, Google announced investments ranging from $180 to $190 billion. With such amounts at stake, the company must demonstrate the profitability of these investments.

Impact on the Stock Market

The announcement of the Frozen v2 chip had a positive effect on investor confidence. Indeed, following the publication of The Information's report, Google's stock rose by about 3% on Monday morning, just before the release of its earnings report. This news thus provided a welcome financial reprieve for the company.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.