Nvidia defended its position at the forefront of the AI hardware market on Tuesday, responding to growing Wall Street concerns that Google’s expanding tensor processing unit lineup could weaken its dominance. The chipmaker argued that its technology remains a generation ahead of competitors, even as major customers explore alternative solutions.
Nvidia Responds to Market Concerns
The company’s comments followed a 3% drop in Nvidia shares, triggered by reports that Meta may partner with Google to deploy TPUs in its data centers. Nvidia said on X that it continues to work successfully with Google, adding that its platform “runs every AI model and does it everywhere computing is done.”
Central to Nvidia’s message was the argument that its GPUs offer broader capability than ASIC chips — specialized processors such as Google’s TPUs that are optimized for narrowly defined tasks. Nvidia highlighted its latest Blackwell architecture as more powerful and flexible compared with single-purpose designs.
Google’s TPU Advances Draw Attention
Google’s in-house chips have gained fresh momentum following the rollout of Gemini 3, a state-of-the-art AI model trained entirely on TPUs. Although Google does not sell the chips directly, it uses them internally and provides cloud-based access for enterprise customers through Google Cloud.
“We are experiencing accelerating demand for both our custom TPUs and Nvidia GPUs,” a Google spokesperson said, signaling the company’s commitment to supporting both technologies.
Nvidia Remains Confident in Its Lead
Analysts estimate that Nvidia still controls over 90% of the AI chip market thanks to widespread adoption of its GPU platforms. CEO Jensen Huang recently acknowledged TPU competition but emphasized that Google remains a buyer of Nvidia hardware, and that its Gemini model can run on Nvidia systems.
Huang also referenced conversations with Google DeepMind CEO Demis Hassabis, noting that AI “scaling laws” — the idea that more compute and more data continue to improve model performance — remain intact. Nvidia argues that these scaling demands will further increase global appetite for its GPUs.