Investors want to know when Nvidia will start booking China revenue again and how strong AI demand will be heading into next year

Nvidia is set to report fiscal second-quarter earnings results on Wednesday.

On Wednesday, Wall Street will be watching Nvidia Corp.’s earnings for information about the health of not only the company, but also the entire stock market and the underlying artificial-intelligence rally.

Going into Nvidia’s fiscal second-quarter report, investors have basically taken for granted that the company will beat its own expectations. The chip designer is expected to report second-quarter revenues of $46.05 billion, above last quarter’s guidance of $45 billion, according to FactSet data. Earnings are expected to come in at $1.01 a share.

Data-center revenues are expected to come in at $41.34 billion, up 57% from a year before. While that’s a strong growth rate, it’s down from the 73% clip that Nvidia posted in its most recent earnings report.

The state of the U.S.-China AI arms race

According to Piper Sandler’s Harsh Kumar, Nvidia’s guidance surrounding its China business will be the focal point of Wednesday’s report.

Nvidia’s chips have been caught in the middle of a technological arms race. President Donald Trump has announced that the U.S. will receive a 15% cut of Nvidia’s China chip sales, while the Chinese government has urged local companies to avoid Nvidia’s H20 chips. In response, Nvidia has reportedly halted production on its H20 chips.

Previous policies prevented Nvidia from doing business in China during the quarter that just completed, and Nvidia is expected to guide ex-China for the coming quarter as well.

However, analysts still see upside potential for H20 chips and will be looking for any clarifying commentary on China sales. In a Monday note, JPMorgan analysts wrote that Nvidia’s existing H20 inventory could bring up to $6 billion in additional revenues. Kumar said the China business could see double-digit growth in the coming quarters.

There could also be indirect ways of tapping into the Chinese market, according to William Blair analyst Sebastien Naji. Chinese companies could gain access to advanced Nvidia chips by renting from hyperscalers based in Southeast Asia, Naji wrote on Monday.

As China seeks to strengthen its domestic chip industry, Nvidia’s ability to extract revenue from China will be a key indicator of just how much of a technological edge the U.S. has.

Supply and demand dynamics

Demand for Nvidia’s chips has outpaced supply for several quarters, and Wall Street will be watching for signs of continued strength to gauge the company’s potential for future earnings beats as well as the growth of the data center industry.

Naji has observed an increase in announced 2026 data center projects, which could potentially add up to $100 billion in revenues for Nvidia in the next year. Demand is increasingly coming from sovereign and neocloud developments outside of the U.S., indicating a widening breadth of business for Nvidia.

The company’s Blackwell chip has enjoyed a robust ramp-up, and the next-generation Rubin chip is scheduled to launch in the second half of next year. Cantor Fitzgerald’s C.J. Muse will be looking for more specific demand projections on Wednesday. He said that Nvidia Chief Executive Jensen Huang should have “strong visibility” into data-center demand into calendar year 2026, including the early ramp of Rubin that’s expected to take place late next calendar year.

According to Kumar, demand could also be heavily influenced by the future of the China business, as any positive developments on that front would result in high levels of pent-up demand from Chinese customers.

Will hyperscalers keep spending?

Hyperscalers, or the companies building out AI infrastructure, have consistently raised their capital expenditure projections and snatched up Nvidia’s graphics processing units.

These companies’ multibillion-dollar budgets are a boon for Nvidia. But last week’s stock-market selloff, led by Big Tech names, made some investors uneasy about the sustainability of these large investments.

Additionally, for Nvidia’s fiscal first-quarter report, the company reported receiving 30% of its revenues from just two customers, indicating high customer concentration.

Kumar remains optimistic about hyperscaler spending, as he believes they “are racing towards providing their users with AGI,” or artificial general intelligence, “and have a game plan to spend to that level, which should last for a few more years in our view.”