Tech giants' AI spending soars, so why is Nvidia still falling?
```
The intensive increase in AI capital expenditure by tech giants should logically be a major boon for Nvidia, but the market has given a completely opposite answer.
On Thursday, Nvidia's stock price dropped more than 4%, falling below the $200 mark, with a single-day drop of nearly $10.
The previous night, Meta, Alphabet, Microsoft, and Amazon successively released their earnings reports, and the four hyperscale cloud providers are expected to invest as much as $725 billion by 2026 to build AI infrastructure.
Nvidia holds about a 90% share of the AI accelerator chip market. Normally, this wave of investment should directly benefit this chip giant. But what investors worry about is that when Nvidia's most important customers begin to develop their own chips on a large scale, its market dominance will face challenges.
Previously, Alphabet announced it would sell its self-developed TPU chips to external customers, and Amazon also emphasized the rapid growth of its self-developed chip business in its earnings call.
Google TPU "breaks out", touching the market’s most sensitive nerve
Alphabet announced that it will sell its self-developed TPU chips to selected external customers, who can deploy them in their own data center infrastructure.
Before, TPUs almost completely served Google's internal ecosystem. Once commercialized externally, TPUs will evolve from a potential competitor to Nvidia’s GPUs into a real market threat.
Although TPUs are generally considered less flexible than Nvidia's solutions for general purposes, they are particularly cost-effective for specific AI application scenarios.
Amazon also emphasized the growth momentum of its self-developed chip business during its earnings call.
According to Bloomberg, Amazon CEO Andy Jassy said the company’s chip business has annualized revenue exceeding $2 billion, posting triple-digit year-on-year growth, with Trainium, its self-developed chip, as the core product.
Wall Street analysts: The wave of self-developed chips poses a "major risk"
For this competitive situation, Wall Street analysts have already issued clear warnings.
Seaport Research semiconductor analyst Jay Goldberg stated bluntly:
This could fundamentally disrupt Nvidia. I think this is quite a significant risk.
Goldberg’s logic is based on the scale and capital strength of the hyperscale cloud providers. These companies are not only Nvidia’s largest customers, but are also continuously investing to become its competitors.
In the first quarter, Alphabet's cloud revenue rose 63% year-on-year to $20.03 billion, with backlog almost doubling to over $460 billion, a considerable portion of which is undertaken by internal TPUs.
At the same time that Meta raised its 2026 capital expenditure guidance to $125-$145 billion, it is simultaneously advancing its MTIA self-developed chip project.
However, not all analysts agree with this pessimistic view.
Stacy Rasgon of Bernstein Research offers a rebuttal, arguing, "Worrying about who wins and who loses is the wrong question." His logic is: the rise of AI agents is causing an explosive growth in computing demand, and the key constraint on the industry now is supply, not demand.
In this context, all chip manufacturers with credible production capacity, including Nvidia, can achieve full sales of their output.
Nvidia currently holds $95.2 billion in supply commitments, with top clients including OpenAI, Anthropic, CoreWeave, and Meta.
Risk warning and disclaimerThe market has risks, and investment should be done cautiously. This article does not constitute individual investment advice and does not consider the investment objectives, financial situation, or needs of any particular user. Users should consider whether any opinion, viewpoint, or conclusion in this article is suitable for their specific situation. Investment based on this is at your own risk. ```