Nvidia stock is 5% below its record: why September 10 suddenly matters

Nvidia stock is 5% below its record: why September 10 suddenly matters
Devesh Kumar
09 Sept 2026, 10:18 AM

powered by

Invezz
NVDA into Sep 10

Buy NASDAQ:NVDA. The stock is near highs but still “cheap” versus compute peers on 2028 earnings, and Sep 10 is a clear catalyst: Huang’s Goldman fireside chat can push AI demand and next-hardware confidence higher, keeping the market willing to pay up for accelerating earnings. This is a momentum + valuation support setup.

Key Risk: Huang fails to lift growth expectations (enterprise adoption and next-platform demand sound flat), so the “cheap” multiple compresses fast.

Hugging Face deal as a hedge

Buy NASDAQ:NVDA with a focus on the $12.93B Hugging Face angle. If Huang reinforces that Nvidia can broaden beyond hyperscalers via open-weight models and enterprise distribution, the market will price in a more durable revenue stream and less customer concentration risk—secondarily supporting NVDA’s long-term margins and stickiness.

Key Risk: The Hugging Face strategy doesn’t translate into measurable enterprise traction (or the deal faces regulatory/implementation delays), so concentration risk stays unresolved.

  • Nvidia sits under 5% below its record before Huang's Goldman fireside chat.
  • Cantor keeps a $350 target as it calls Nvidia the cheapest compute stock.
  • Customer concentration and custom chips remain key risks to Nvidia's rally.

Nvidia stock NASDAQ:NVDA is sitting less than 5% below its record high, but the next catalyst arrives on September 10.

Chief executive Jensen Huang will participate in a fireside chat at Goldman Sachs’ Communacopia + Technology Conference at 8:50 a.m. PT, where investors will listen for signals on AI demand, enterprise adoption and Nvidia’s next hardware.

Nvidia closed at $225.73 on Tuesday, down 2.01%, leaving the stock about 4.6% below its $236.54 all-time high.

Nvidia is near a record, yet analysts still call it cheap

Nvidia’s shares are close to a record after another AI-driven rally, yet some analysts argue earnings expectations are rising even faster than the stock.

Cantor Fitzgerald analyst C.J. Muse reiterated an Overweight rating and a $350 price target on September 8. Muse said Nvidia trades at the “cheapest” valuation among compute names based on calendar-2028 earnings estimates.

Cantor also argued that Nvidia remains under-owned by hedge funds and long-only managers.

That makes Huang’s appearance more important.

If he reinforces expectations for accelerating revenue or a widening customer base, investors may continue to view Nvidia as inexpensive relative to future earnings.

But that support depends on estimates continuing to rise. If growth expectations flatten, the stock becomes harder to defend.

Enterprise demand is becoming the next big test

Nvidia’s next leg depends on proving AI demand is expanding beyond a small group of hyperscalers.

Recent results from Dell strengthened that argument. Dell raised its annual outlook after reporting record revenue and a large AI-server backlog.

D.A. Davidson analyst Gil Luria told MarketWatch that Dell’s results were another sign the enterprise AI-compute market has momentum “beyond the current hyperscaler market.”

They will listen for commentary on enterprise adoption, sovereign AI, supply constraints and Feynman, Nvidia’s next architecture.

A broader customer base would make Nvidia’s growth story more durable.

If AI spending remains concentrated among Microsoft, Amazon, Meta and other technology giants, investors will keep worrying about concentration.

Customer concentration remains the uncomfortable risk

The bullish case has a major weakness, as Nvidia’s latest regulatory filing showed that three direct customers represented 16%, 15% and 13% of total revenue in the first half of fiscal 2027.

Investor Dan Niles highlighted the same issue this week, noting that Nvidia’s largest customers are increasingly designing their own application-specific chips.

That creates an unusual tension: the companies funding Nvidia’s growth also have the strongest incentives to reduce their dependence on its GPUs.

Nvidia’s planned $12.93 billion acquisition of Hugging Face could help address that risk by giving the company access to more than 18 million developers and a stronger route into enterprise AI.

Niles said that open-weight models could eventually dominate LLM usage, allowing Nvidia to sell a broader stack directly to enterprises rather than relying heavily on hyperscalers.