[ad_1]
Want more stock market and economic analysis from Phil Rosen directly in your inbox? Subscribe to Opening Bell Daily’s newsletter.
The market is acting like Google is coming for Nvidia.
Alphabet’s vertical integration with its new Gemini 3 AI model and its custom AI chips has fueled enthusiasm for its standing in the AI race while spooking Nvidia shareholders, creating a widening performance gap between the two stocks.
Specifically, worries seem to be rising that the customers who have long relied on Nvidia chips could soon turn to Google. Indeed, The Information reported this week that Meta is in talks to use Google’s Tensor Processing Units (TPUs).
That, in theory, redirects billions in business from Meta out of Nvidia’s pocket and into Google’s.
The report alone was enough to wipe out some hundreds of billion in Nvidia’s market cap on Tuesday while boosting Alphabet’s, shrinking the valuation gap between the two heavyweights to its narrowest since April.
To be clear, Nvidia’s graphics processing units (GPUs) remain the gold standard for the AI industry. But Google’s TPUs — which power its highly praised Gemini 3 — are cheaper to develop and require less power.
Some industry experts estimate that TPUs offer up to four times better performance per dollar than comparable GPUs.
So while the technology itself may not be apples-to-apples competitive, the economics of choosing one over the other does seem to be a hit against Nvidia, in the market’s view.
Nvidia, for its part, seemed to brush off the news entirely.
“We’re delighted by Google’s success,” Nvidia’s communications team wrote in a statement. “They’ve made great advances in AI and we continue to supply to Google. Nvidia is a generation ahead of the industry — it’s the only platform that runs every AI model and does it everywhere computing is done.”
Shares of Alphabet have more than doubled the returns of Nvidia so far this year, though only in recent weeks have technologists seemed to concede that Google could be winning the AI race.
Its outperformance is underscored by its unique “full stack” advantage:
- Google designs its own chips
- Google trains its own AI models
- Google has its own distribution channels
That’s left companies like Oracle, which bought billions of dollars worth of Nvidia chips to rent out, lagging the market as it reprices to a landscape that includes Google’s more economical alternatives.
While both companies are sure to compete for years to come, these developments confirm that the AI chip battle is no longer a monopoly.
The final deadline for the 2026 Inc. Regionals Awards is Friday, December 12, at 11:59 p.m. PT. Apply now.
[ad_2]
Phil Rosen
Source link