Upgrade to SI Premium - Free Trial

Nvidia gives back big early gains despite Huang dismissing AI bubble concerns

November 19, 2025 4:32 PM
(Updated - November 20, 2025 1:19 PM EST)

Investing.com -- Nvidia stock gave back its significant early gains and more, despite providing strong guidance for the current quarter and fiscal third-quarter results that exceeded Wall Street estimates Wednesday afternoon as the AI arms race continued to drum up demand for its latest AI Blackwell chips.

After trading up over 5% earlier during Thursday’s session, NVIDIA Corporation (NASDAQ: NVDA) traded down 1.7% at the 1 PM hour on Wall Street. The sell-off came despite news that CEO Jensen Huang tried to assuage investor fears of an "AI bubble." Some bears pointed to rising accounts receivables, although bulls said that the increase corresponds to the surge in revenue.

For the three months ended Oct. 26, the company reported Q3 adjusted earnings per share of $1.30, up from $0.78 in the same period a year earlier, on revenue of $57.01B, up 62% from a year ago. Analysts polled by Investing.com anticipated per-share income of $1.25 and revenue of $54.8B.

The company’s data center business, which makes up the bulk of revenue and includes the next-generation Blackwell AI chips as well as prior-generation Hopper AI chips, saw revenue climb 66% to $51.22B, beating estimates of $49.09B.

CFO Colette Kress said Blackwell momentum is driving unprecedented visibility, reiterating that Nvidia currently has “visibility on $500 billion in revenue from the beginning of the year to the end of 2026,” with further opportunities on top of that. She added that roughly half of Nvidia’s longer-term opportunity will come from hyperscalers transitioning to accelerated computing and generative AI.

Blackwell sales were fueled by hyperscalers including Microsoft Corporation (NASDAQ: MSFT), Amazon.com Inc (NASDAQ: AMZN), and Alphabet Inc Class A (NASDAQ: GOOGL), with Huang saying demand is “off the charts” and cloud GPUs are effectively sold out. Kress later noted that the GB300 chip accounted for about two-thirds of total Blackwell revenue in the quarter.

"Amidst a wide range of growing AI concerns, NVDA answered the bell with GB300 shipments driving healthy upside to estimates," Jefferies analyst Blayne Curtis commented in a note. "These results and commentary should help steady the ship for the AI trade into the end of the year."

Jefferies lifted its price target on the stock to $250 from $240.

Separately, Melius Research analyst Ben Reitzes also raised the target price to $320 from $300 after the report. He highlighted Nvidia’s "revenue acceleration to 65%+ with this kind of margin at scale, benefiting from the greatest paradigm shift of my lifetime is worth more than ~25x next year’s EPS."

Networking, which also forms part of the data center business and which Nvidia has touted as a growth engine, saw sales rise 162% from a year ago.

The gaming business, meanwhile, reported a 30% rise in sales to $4.27B. Gross margins fell 1.4 bps to 73.6%, though beat estimates of 73.4%.

China, geopolitics, and chip mix

Kress said sales of Nvidia’s new H20 chip specialized for China totaled about $50M in Q3, while tougher U.S. export rules resulted in “sizable purchase orders” for Hopper-generation products failing to materialize. She added the company was “disappointed” by the new restrictions but remains committed to engaging with regulators.

CEO pushes back on AI bubble narrative

Huang directly addressed growing debate about an AI spending bubble, saying Nvidia is seeing “something different,” driven by three concurrent platform shifts:

  1. The move from CPU general-purpose computing to GPU-accelerated computing.
  2. AI reaching a tipping point, with generative AI replacing classical machine learning.
  3. The transition to agentic AI, which he said most industries haven’t even engaged with yet.

“Don’t just look at hyperscalers… Look at the world. Look at different industries,” Huang said, adding that Nvidia’s inference leadership and the number of customers switching to Nvidia platforms “after exploring others” is increasing.

Huang also emphasized Nvidia’s central position in the ecosystem, noting its platform runs “every major AI model,” including OpenAI, Anthropic, xAI, Gemini, and others. He pushed back on "circular investment" criticism, specifically involving OpenAI, saying the move to invest in the ChatGPT-maker is intended “to expand our ecosystem and support their growth.”

Guidance beats

Looking ahead, Nvidia said it expected about $65B in fiscal fourth-quarter revenue, plus or minus 2%, beating analyst estimates of about $61.84B. Kress reiterated that Nvidia’s revenue visibility through 2026 exceeds $500B, with room for further upside as more industries adopt agentic AI.

Categories

Investing

Next Articles