Nvidia has a lot of questions to answer, and none of them are about demand for its AI chips

Nvidia’s earnings report comes at a tenuous time for the company and the AI trade as a whole.

Fund managers think companies (read: hyperscalers) are investing too much. The good news about demand is known: Nvidia effectively preannounced its revenue outlook when CEO Jensen Huang touted more than $500 billion in orders for its flagship chips through calendar year 2026. And shares recently traded near the bottom of the $180 to $210 range they’ve been oscillating around since the end of September.

“NVDA near-term is facing the tough task of meeting high (earnings) expectations AND high skepticism around AI capex, likely only resolved when broader market volatility (shutdown, interest-rates) subsides,” Bank of America analyst Vivek Arya wrote on November 14, while upgrading his earnings estimates for this year as well as the following two. “We look for management to provide reassurance around demand and supply and believe muted sentiment (stock -10% since GTC order raise) a contrarian positive heading into the print.”

The company’s had a busy two days heading into its earnings report, announcing a partnership with Microsoft and Anthropic as well as investing in Brookfield’s new AI infrastructure fund, both of which are poised to drive more demand for its GPUs.

Expectations and reality

Analysts are expecting the company to deliver $55.2 billion in sales in its fiscal Q3 2026, with adjusted earnings per share of $1.26. That roughly $8.4 billion in revenue growth from Q2 to Q3 would be more than the total sales generated by over 370 S&P 500 companies in their most recent quarter.

But the most valuable company in the world has earned that moniker despite, rather than because of, how investors have reacted to its quarterly reports as of late. Just once over its past five reporting periods has the chip designer gained in the week following earnings.

This time, there should be no questions about the appetite for Nvidia’s AI chips. What’s in doubt is how much supply chain snarls might impact its ability to meet that colossal demand, whether margins might face some pressure along the way, and if competition will eat away at its dominant market position over time.

Nvidia’s issue is that too many companies in the AI boom aren’t Nvidia

To this end, Morgan Stanley analyst Joseph Moore recently flagged one oft overlooked aspect of the AI boom: that it came out of nowhere, at a time when there was excess capacity in semiconductor production. That’s no longer the case.

If Nvidia’s GPUs are the brains of AI, you still need a host of other chips to serve as the supporting elements of the nervous system. On that note, high-bandwidth memory prices have been surging as supply remains ultra-tight, spurring huge gains for the likes of Micron.

“Growing is going to require dynamic supply chain management, aggressive willingness to commit to take-or-pay volumes, and, likely, higher input costs for wafers and DRAM,” Moore wrote.

Nvidia is better positioned than Broadcom or AMD to grapple with this shift, per Moore, but even the heavyweight may not be immune from some diminished profitability.

“Does that mean margin erosion for all three players?” he added. “It might, as higher input costs — especially HBM DRAM — might be tough to fully pass on, but pricing power is pretty high.”

JPMorgan analyst Harlan Sur said that Nvidia’s supply chain partners “have demonstrated strong execution” to date in ramping Blackwell and Blackwell Ultra shipments over the past three to four months, and he expects the company to deliver better-than-expected results, along with strong guidance.

How good does the print have to be?

That being said, he cautions that this may not be enough for investors to cheer the results:

“We still see supply chain capacity as a gating factor to revenue growth for NVDA well into calendar year 2026. We consequently expect the stock to key more to management’s framing of the trajectory for the Blackwell/Blackwell Ultra ramp into fiscal 1H27 (C1H26), and the manner in which questions around investors’ key concerns are addressed, including the sustainability of growth in AI spending (the JPM global team concluded in a recent deep dive that funding sources will be ample through 2030), the impact of power constraints on DC infrastructure rollouts given an estimated ~120 GW of capacity slated to come online over the next five years (current lead times for new natural gas turbines have ballooned to 3-4 years, and nuclear plants have historically taken 10+ years to build), and the effect (if any) of component cost inflation (memory, wafers, etc.) on gross margins.”

When it comes to how markets will interpret the results, Wedbush Securities analyst Dan Ives takes a more optimistic and straightforward view: the sheer size of the numbers put up by the chip designer will be too impressive to ignore.

“Datapoints from Nvidia this week will be important to convince ‘on the fence investors’ that this AI spending trend is an unparalleled moment in modern tech history and is NOT a bubble moment,” he wrote. “In our opinion Nvidia’s earnings/guidance and general bullish commentary from Jensen will be a positive catalyst for tech stocks into year-end and give an important validation moment around global demand drivers/ magnitude for the AI Revolution from Jensen’s key perch.”