In theory, the numbers coming out of the latest Big Tech earnings reports would seem like good news for Nvidia. It’s clear that big cloud companies are ramping their spending on artificial-intelligence chips, and Nvidia is the biggest player in that market.
Yet shares of Nvidia fell 4.6% on Thursday, with Wall Street seemingly concerned about heightened competition in the market for AI processors.
Morningstar analyst Brian Colello told MarketWatch that he thinks investors are worried about the recent progress Google has made on its tensor processing units, or TPUs, and that Amazon.com has made on its custom Trainium chips.
“If both hyperscalers, plus Anthropic, are successful in deploying more and more TPUs and Trainium, then it could crowd out the potential opportunities for Nvidia beyond 2027,” he said.
Nvidia’s stock decline on Thursday came as shares of rivals Advanced Micro Devices and Broadcom — the latter of which partners with Google on the TPUs — rose 5.2% and 3%, respectively.
Seaport Research analyst Jay Goldberg told MarketWatch that he thinks investors may be misreading an announcement from Qualcomm, which teased that it ispreparingfor a custom-chip engagement with a hyperscaler later this year. He said investors could be interpreting the announcement as “one more Nvidia competitor entering the fray.”
“I don’t think Qualcomm really is that, but [we’re] certainly seeing a lot of progress on Nvidia alternatives this week,” he said, referring to Google’s TPUs and Amazon’s Trainium chips.
Goldberg also noted that recent news of DeepSeek moving to use Nvidia competitor Huawei’s Ascend 950 AI chips for its V4 artificial-intelligence model could be a factor weighing on the stock.
That said, Nvidia still has plenty of believers.
Tigress Financial Partners analyst Ivan Feinseth told MarketWatch that there’s still ample room for Nvidia to cash in on robust data-center demand and benefit from the increasing capital-expenditure budgets mentioned this week by Alphabet, Meta Platforms and Microsoft.