Alphabet shares rise as Google reportedly develops artificial intelligence chips to improve Gemini efficiency
On Monday, reports emerged that Google is developing a new artificial intelligence server chip called “Frozen v2” to make it Gemini Artificial Intelligence Models run more efficiently. The new chip is specially built for Gemini AI. It is designed to increase speed and reduce the computing power required to answer user questions.
Frozen v2 will permanently build part of Gemini’s artificial intelligence architecture directly into the chip. This means the chip can perform many tasks without having to run the same calculations over and over again. Due to this design, the chip will reduce the number of calculations and the movement of data within the system. This will make Gemini faster and more energy efficient.
Google explains AI chip plans
Alphabet did not directly confirm the project, but the company told CNBC that its engineers are constantly researching and testing new technologies to improve performance and efficiency. letter It said not every research project will become a real product, but trying new ideas is an important part of its strategy. The company also says its hardware and software are designed together so its AI systems are highly optimized for real-world workloads.
Google engineers believe Frozen v2 can generate six to 10 times more AI tokens using the same amount of power compared to Google’s latest tensor processing unit (TPU). AI tokens are small text snippets that are read and generated by AI models. Producing more tokens with the same computing power means higher efficiency and lower operating costs. Frozen v2 is not expected to replace Google’s TPU. Instead, it will be a more specialized chip in Google’s lineup of custom AI hardware.
Also read: Microsoft backs AMD Helios AI system to challenge Nvidia’s dominance
Google aims to launch in 2028
According to reports, Google plans to deploy this new chip in 2028. A major goal of the project is to solve Google’s internal AI computing power shortage. The company has been working hard to meet the growing demand for artificial intelligence infrastructure. The computer shortage reportedly created internal pressure and even forced Google Cloud to turn away some outside customer business. Last month, Google reportedly agreed to pay SpaceX nearly $1 billion a month to help boost computing power and fulfill its enterprise artificial intelligence promises, CNBC reported.
Frozen v2 has one important drawback. The chip will only work properly if future Gemini models continue to use the same core architecture. If Google makes significant changes to Gemini’s design in the future, Frozen v2 may become less useful. According to The Information, this is because the chip is built for a specific artificial intelligence architecture. Google is reportedly currently considering Frozen v2 as a test project. The company does not plan to mass-produce its general-purpose TPU.
Google faces artificial intelligence competition
Google’s artificial intelligence business faces other challenges. The next version of Gemini Pro has reportedly been delayed. Google has also lost several senior artificial intelligence researchers to competing companies, adding to pressures in the fast-moving AI race. China’s AI models are becoming more powerful competitors. The report states that they currently account for approximately 45% of AI token usage in U.S. enterprises.
Competition intensified over the weekend after Moonshot AI and Alibaba released new artificial intelligence models, which reportedly narrowed the performance gap with the leaders American artificial intelligence model. Meanwhile, Google DeepMind CEO Demis Hassabis is meeting with U.S. lawmakers in Washington, D.C., to discuss artificial intelligence regulation. According to CNBC, Hassabis proposed establishing a FINRA-style artificial intelligence regulator that would be overseen by the federal government, funded primarily by the artificial intelligence industry, and responsible for testing advanced artificial intelligence models for national security risks before release.
Alphabet invests in artificial intelligence chips
Alphabet has been investing heavily in custom artificial intelligence chips as part of its long-term strategy to reduce its reliance on third-party hardware vendors and build more of its own artificial intelligence infrastructure. More efficient AI chips could help Google lower computing costs, improve Gemini’s performance and solidify its position against rivals such as OpenAI, Anthropic and China’s leading AI companies.
Investors welcomed the report, driving Alphabet shares higher as they believe more powerful in-house AI hardware can support the company’s long-term AI growth and competitiveness.
