By Britney Nguyen
While investors debate which company will win in the AI chip market, a Bernstein analyst says the question is more about whether or not the market is still growing
Nvidia CEO Jensen Huang said in October that the company has visibility into $500 billion in revenue from its Blackwell and Rubin AI platforms through 2026.
Alphabet already passed Microsoft to become the third-largest company by market capitalization, and it’s closing in on Apple. After that, Nvidia would be next in its sights.
The dynamic has two sides. Alphabet’s stock (GOOG) (GOOGL) has been charging higher on enthusiasm for the company’s strides in artificial intelligence. And Nvidia’s stock (NVDA) has fallen well off highs that translated to a $5 trillion valuation as recently as Nov. 3, according to Dow Jones Market Data.
Partly fueling Nvidia’s stock declines, which amount to more than 4% in Tuesday’s session alone, is the idea that Alphabet’s momentum in AI could steal chip business away from Nvidia, the overwhelming market leader. That was underscored late Monday by a report from The Information saying that Meta Platforms (META) is interested in using Google’s tensor processing units in its data centers.
See more: Google is crushing it. Why that’s worrying investors in Nvidia and other AI stocks.
Alphabet’s perceived “AI comeback” with its TPUs and improved Gemini AI models has led some investors to be “petrified that Alphabet will win the AI war,” Melius analyst Ben Reitzes said in a Monday note.
But Bernstein analyst Stacy Rasgon said investors are paying too much attention to the question of whether graphics processing units like Nvidia’s will win out over Alphabet’s TPUs, which are a form of application-specific integrated circuit enabled by Broadcom (AVGO) and custom-built for specific tasks.
Don’t miss: Google may be sitting on a $900 billion gem that could disrupt Nvidia’s dominance
The Google TPUs “are the only really successful ASIC program,” Rasgon wrote, and “certainly the only one that has truly reached scale.” Nonetheless, the “overarching theme” for the industry is that there is a scarcity of compute power.
“If anything,” Meta’s potential turn to Google for TPUs would imply an effort to secure additional compute capacity, Rasgon said. Meta has already indicated that it expects its capital-expenditure growth to be “noticeably larger” in 2026 than it will be in 2025, as the company boosts its AI efforts.
Rasgon pointed to other AI players such as OpenAI “taking a similar diversified approach” to secure chips from Advanced Micro Devices (AMD) and Broadcom – “something that has not slowed [Nvidia] down in the slightest given the state of compute demand.” He also noted that Nvidia CEO Jensen Huang disclosed last month that it had visibility into $500 billion in revenue from its Blackwell and Rubin AI platforms through 2026, some of which has already shipped.
Therefore, “the question of ‘ASIC or GPU’ kind of misses the point,” in Rasgon’s view.
“Right now the real question should really be, ‘Is the opportunity in front of us still big, or is it not?'” he wrote. The AI hardware market has likely not yet reached “a mature, saturated” stage, Rasgon said, and the size of the market, rather than the share, is what matters.
“If it’s big, both GPU and ASIC should thrive,” Rasgon said. If not, “they’re both in trouble.”
In a statement shared on X, Nvidia said the company is “delighted by Google’s success,” and that it plans to continue supplying the tech giant with its chips.
However, the chip maker noted that it “is a generation ahead of the industry,” and has “the only platform that runs every AI model and does it everywhere computing is done.”
“Nvidia offers greater performance, versatility and fungibility than ASICs, which are designed for specific AI frameworks or functions,” the company said.
Mizuho desk-based analyst Jordan Klein acknowledged the recent “winner-take-all” sentiment in the AI trade in a Tuesday note – a vibe he said is “gaining momentum” thanks in part to the positive reception of Google’s Gemini 3.
But “the AI arms race will not be won or lost” this month, Klein wrote. Rather, he senses “a MARATHON that will see many different lead changes.”
“The biggest players with frontier models will constantly spend and invest to gain an advantage and edge over the field,” Klein said, which will lead to “more investment, hiring and power/compute/memory/high-speed connectivity buildouts and expansion.”
For now, the AI trade is not looking at “a winner-take-all situation,” he added.
-Britney Nguyen
This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
11-25-25 1321ET
Copyright (c) 2025 Dow Jones & Company, Inc.
