Nvidia earnings could expose cracks in the AI boom

After the conclusion of trading on Wednesday, August 26, Nvidia will announce the results of its second quarter, and these numbers will be significant for more than just its own share price. The results will also be monitored by a variety of chipmakers, cloud providers, and suppliers of equipment throughout the entire AI supply chain in order to try to assess whether the increased spending on the global infrastructure boom translates into actual orders for Nvidia’s machinery.

According to the July announcement from Nvidia’s investor relations team, the call will take place at 2 p.m. Pacific Time, and the results will be made public around 1:20 p.m.

Why does the whole AI market key off one report

Very few announcements about profits have this much importance for a whole sector. Chips from Nvidia are present in the data centers that are being developed by Microsoft, Alphabet, Amazon, Meta, and Oracle, which means that the sales figures of the company provide real-time information about how fast the expansion is taking place.

In August, TrendForce upgraded its AI server shipment prediction for 2026 to almost 31% compared to the previous year, as large cloud providers boosted their orders for Nvidia’s vision of rack-scale infrastructure. According to the market research firm, the amount spent by the eight largest cloud companies this year will go up by about 90% to over $886.7 billion. IDC confirmed that the same trend is presently being observed, with its data showing that global server spending has increased by 30.7% in Q1 of 2026, due to the introduction of GPU technology.

What Nvidia’s earnings are actually made of

For those unfamiliar with Nvidia’s statistics, the bottom line is as follows: hardware for data centers is the most important part of their business operations.

When Nvidia last reported on May 20, data center sales reached a record $75.2 billion out of $81.6 billion in total revenue. The remaining $6.4 billion came from Edge Computing, Nvidia’s category for PCs, consoles, robotics, and automotive chips.

Furthermore, the company has decided to separate its data center reporting into two divisions: Hyperscale, which will include public clouds and the biggest internet businesses, and the second, which will only serve AI clouds as well as industrial and commercial markets. Revenues for the data center increased by 85 percent year-on-year in the first quarter, while the company projected total revenues of around 91 billion dollars for the quarter about to finish.

The spending question hangs over the numbers

A major question as we approach Wednesday is whether the budgets for cloud computing can continue to grow at this pace.

According to a Reuters report released in July, major hyperscalers will spend more on capital expenditures than they are able to generate in free cash flow by 2027, with capex shooting up by approximately $534 billion, compared to a lesser rise in operating cash.

“Investors are underestimating how fundamentally AI is changing the Big Tech business model,” Futurum Equities strategist Shay Boloor told Reuters, arguing that AI is turning asset-light software companies traditionally into infrastructure-heavy businesses.

Those companies are also Nvidia’s largest clients. When they start to cut back on spending, the repercussions will quickly fall upon Nvidia’s order volume.

The China gap Nvidia has already priced out

Nvidia’s forecast excludes one very important industry.

The company’s roughly $91 billion outlook assumes no data center compute sales to China, reflecting how much ground Nvidia has lost there. Cryptopolitan reported that Nvidia’s share of China’s AI chip market could fall to about 8% in 2026 from nearly 40% a year earlier, according to a Bernstein estimate, as Huawei and other domestic suppliers move above half the market.

Chief executive Jensen Huang has said Nvidia “largely conceded that market” after years of tightening US export restrictions. Small batches of H200 accelerators have begun returning under case-by-case licenses, but Chinese companies are increasingly directing their budgets toward domestic chips.

Who actually captures the economic returns?

One additional current wrinkle: Nvidia is reportedly telling major customers that AI-server prices could rise more than 15% from early 2027, largely because of surging memory costs. That could become important for gross margins and the economics of the entire AI infrastructure buildout.

Reuters reported last week that investors are increasingly shifting the question from “How much are Big Tech companies spending?” to “Who actually captures the economic returns?” Strong cloud growth and persistent capacity constraints have so far helped ease some concerns about AI capex. Nvidia’s real test is no longer whether AI demand exists. It’s whether $700B-plus of hyperscaler spending can keep translating into accelerating chip revenue.

Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.