PANews March 2 News, "Wood Sister" Cathie Wood's ARK Invest investment management company recently predicted that NVIDIA will face increasingly fierce competition in the coming years. By 2030, custom AI chips will account for more than one-third of the computing market. Meanwhile, Google is also continuing to position its Tensor Processing Units (TPU) as an alternative to NVIDIA GPUs, and there are reports that Meta has agreed to lease TPUs for advanced AI development.

BlockBeats News, February 27 — According to The Information, Google has reached a multi-billion dollar deal with Meta for artificial intelligence chips. Meta has signed an agreement to lease Google's AI chips—specifically TPU—for developing new AI models. Meta is also in talks with Google, planning to purchase TPUs for its data centers as early as next year. Google has established a joint venture aimed at leasing its tensor processing units to other AI clients.

Odaily Planet Daily reports that AetherMind, an intelligent network dedicated to building a decentralized resource ecosystem, announced that token subscription began on February 10th and will continue for three days. The project will deeply integrate global GPU/TPU computing power through smart contracts, and utilize zero-knowledge proofs and cross-chain interoperability to create a transparent and secure technical foundation for AI inference, training, and data trading, thereby providing seamless and efficient AI services for every DApp.

According to a report from Jinse Finance and The Information, Google is now promoting its Tensor Processing Units (TPUs) to major clients, including Meta Platforms, for use in their own data centers, which directly undermines Nvidia's core business model.

According to reports from Jinse Finance and The Information, Google is actively expanding its TPU market, including reaching a protocol with cloud service provider Fluidstack: if Fluidstack fails to pay the rent for the new data center in New York, Google will provide a guarantee of up to $3.2 billion.

Jin10 reports that there has been an increasing buzz in the market about Google's (GOOG.O) self-developed TPU, raising concerns about its impact on NVIDIA's market share in AI chips. NVIDIA CEO Jen-Hsun Huang stated that the AI market is vast and growing rapidly, with competition occurring daily, and NVIDIA must "run faster." He emphasized that NVIDIA's positioning is "very unique and solid," with a core focus on the high versatility of its GPUs and platform, which can be deployed in major cloud environments and is currently the only system capable of running all AI models. Regardless of customer needs, NVIDIA can provide complete solutions.

According to a report by China Merchants Securities International, Google Cloud has secured a NATO AI cloud contract, which will provide a dedicated secure cloud system and utilize AI to enhance operational efficiency. At the same time, Google is competing for TPU orders with Meta, showcasing its market opportunities. Investor concerns about Google may ease, maintaining a "buy" rating with a target price raised to $390.

Golden Finance reported that Nvidia stated on Tuesday that its technology remains a generation ahead of the industry in response to Wall Street's concerns about Google's AI chips potentially threatening Nvidia's dominance in AI infrastructure. Nvidia said, "We are pleased with Google's success—they have made significant progress in the AI field, and we will continue to provide products to Google. Nvidia currently leads the industry by a generation—we are the only platform that can run all AI models and is versatile across various computing scenarios." At the time of this statement, Nvidia's stock fell 3% on Tuesday, following reports that one of its key clients, Meta, may reach an agreement with Google to use Google's Tensor Processing Units (TPU) in its data centers. Nvidia stated in the post that its chips are more flexible and powerful compared to so-called ASIC chips (like Google's TPU), which are typically designed for a single company or single function.

In pre-market trading on Tuesday, Nvidia's stock price fell by 3.2%, following reports that Meta is in talks to use Google's AI chips. Boosted by this news, Alphabet's stock price rose by 2.1%, as investors digest the potential shift in the AI hardware landscape. The Information reported on Monday that Meta is considering deploying Google's Tensor Processing Units (TPUs) in its data centers before 2027. The social media giant may rent TPUs from Google Cloud as early as next year.

Google's parent company Alphabet is in talks with companies like Meta to open up the rights to use its self-developed Tensor AI chips, expanding competition with Nvidia. Meta is considering purchasing Google's TPU chips and renting Computing Power, which could pressure Nvidia and AMD.