Key Points
- Nvidia and AMD stocks fell 4% on Tuesday morning. Meta is considering buying billions of AI chips from Google.
- The deal could see Google TPUs in Meta data centers by 2027.
- SoftBank shares dropped 10% due to fears of rising competition.
- Google and Broadcom stocks rose as investors cheered the news.
Nvidia and AMD took a hit on Tuesday morning, with both companies seeing their stock prices drop by 4% before the market even opened. The decline follows reports that tech giant Meta Platforms is in serious talks to buy billions of dollars’ worth of artificial intelligence chips from Google, rather than relying solely on its traditional suppliers.
According to a report by The Information, Google is actively pitching its custom “Tensor Processing Units” (TPUs) to major clients. This list includes financial institutions and, most notably, Meta. Currently, Meta builds its massive AI infrastructure primarily using Nvidia’s graphics processors.
However, the report suggests that Meta is considering a multi-billion-dollar investment to install Google’s chips in its own data centers by 2027. Additionally, Meta might start renting access to these chips through Google Cloud as early as next year.
This potential change in strategy shook the semiconductor industry. Investors worry that if big companies like Meta switch to Google’s hardware, Nvidia’s absolute dominance in the AI market could weaken. The news also hurt SoftBank Group, which saw its shares tumble 10%. SoftBank is a major backer of OpenAI, and a stronger Google ecosystem could threaten its investment.
While Nvidia and AMD suffered, the news was good for Google and its manufacturing partner, Broadcom. Both companies saw their stock prices rise in pre-market trading. Investors appear optimistic about Google’s ability to compete in the hardware space. Broadcom, which helps design and build the TPUs, stands to gain significantly if Google secures these massive contracts.
The market reaction highlights how fierce the competition for AI chips has become. For years, Nvidia was the only real option for high-performance AI tasks.
Now, with companies like Google offering viable alternatives and big clients looking to diversify, the landscape is shifting. The race is no longer just about who has the best software, but who controls the physical silicon that powers it.
