Nvidia’s (NVDA) stock is down 4% after media reports surfaced that Meta Platforms (META) plans to use microchips designed by Google parent company Alphabet (GOOGL).
News site The Information was first to report that Meta is considering using Alphabet’s tensor processing unit (TPU) microchips in its artificial intelligence (A.I.) data centers starting in 2027.
Meta may also rent TPUs from Alphabet’s cloud computing unit as earlier as next year as it races to stay ahead in the global A.I. industry.
Alphabet launched its first-generation TPU in 2018 and it was initially designed for its own internal use with its cloud computing business.
However, Alphabet has now launched more advanced versions of its signature microchip that are designed to handle A.I. workloads and is selling them to other companies like Meta.
TPUs are a customizable chip and experts say this gives Alphabet an advantage over rivals as it can offer customers a highly efficient product for A.I applications and models.
If Meta uses the TPUs, it would be big win for Alphabet and help validate the search engine giant’s microchip technology.
For Nvidia, Alphabet’s TPUs represent a competitive threat.
Nvidia remains the market leader with its graphics processing unit (GPU) microchips that have become the main piece of hardware underpinning A.I. infrastructure worldwide.
However, Nvidia is facing growing threats to its A.I. microchip dominance from Alphabet, as well as Advanced Micro Devices (AMD), Arm Holdings (ARM), and others.
Companies that are rapidly building A.I. infrastructure, such as Microsoft (MSFT) and OpenAI, are searching for a more diversified supply of microchips to reduce their reliance on Nvidia.
The competition in the chip space and decline in Nvidia’s share price come as a debate rages about whether there is an A.I. bubble amid high tech company valuations.
Prior to today (Nov. 25), NVDA stock had risen 32% this year to trade at $182.55 U.S. per share.
