Nvidia Reports Second-Quarter Fiscal 2027 on Wednesday Against Its Own $91.0 Billion Guidance, With the October-Quarter Outlook the Real Test for the Chip Complex

by Karl Lindström · August 24, 2026

Semiconductors & Hardware Reporter, Frontier Tech Wire

Nvidia reports second-quarter fiscal 2027 results on Wednesday. The company said in a July 29 notice that it would hold its conference call on Aug. 26 at 2 p.m. Pacific time, 5 p.m. Eastern, covering the quarter that ended July 26, 2026, and that written commentary from chief financial officer Colette Kress would be posted to its investor site immediately after results are announced, at approximately 1:20 p.m. Pacific. Nothing has been reported yet; everything below is the bar, not the outcome.

The bar was set on May 20, when Nvidia guided second-quarter revenue to $91.0 billion, plus or minus 2%. On the same guidance the company put GAAP gross margin at 74.9% and non-GAAP gross margin at 75.0%, each plus or minus 50 basis points, with GAAP operating expenses of about $8.5 billion and non-GAAP operating expenses of about $8.3 billion. One line in that outlook has done a lot of work in the months since: Nvidia said it was not assuming any Data Center compute revenue from China.

The first quarter is the base those figures build on. Nvidia reported revenue of $81.6 billion for the three months to April, up 20% sequentially and 85% year over year. Data Center revenue was $75.2 billion, up 21% sequentially and 92% year over year, split between $60.4 billion of compute — up 77% year over year and 18% sequentially — and $14.8 billion of networking, up 199% year over year and 35% sequentially. That networking growth rate is the number the optical and interconnect supply chain has been trading off, and it is worth watching whether the second quarter sustains it.

On profitability, Nvidia reported first-quarter GAAP diluted earnings per share of $2.39 and non-GAAP diluted EPS of $1.87. The GAAP figure sitting above the non-GAAP figure is unusual and readers should treat the two as separate measures rather than interchangeable ones; the company's release is the place to reconcile them. GAAP gross margin came in at 74.9% and non-GAAP at 75.0%.

Third-party estimates circulating ahead of the print sit above the company's own guide, but they are not a single agreed number. An Investing.com preview published Aug. 21 cited consensus of $91.9 billion in revenue and $2.08 in EPS for the July quarter, and put Benchmark's forecast at $92 billion of revenue and $2.10 of EPS, which it characterised as slightly above consensus. That preview did not name the firm that compiled the consensus it quoted. Different compilers publish materially different numbers for the same Nvidia quarter, so $91.9 billion should be read as one preview's figure on one date, not as an agreed market number. The only revenue figure in this article that comes from Nvidia is its own $91.0 billion plus or minus 2% guide.

What comes after the print is the October-quarter outlook, and Nvidia has not issued one — it is due to do so on Wednesday. The same Investing.com preview, dated Aug. 21, cited a third-quarter consensus of $103.96 billion in revenue and $2.37 in EPS, and said that if Nvidia guides below that level "the unwind could be swift," pointing to the stock's 2.22 beta. That figure carries the same caveat as the July-quarter one: it is a forward expectation attributed to an unnamed compiler by a single preview on a single date, not a company disclosure and not a fact about the future. Sequentially, roughly $104 billion against a $91.0 billion guided quarter would imply growth continuing at a pace that requires the AI factory buildout to keep absorbing supply at current rates.

Two variables carry the read-through to the rest of the chip complex. The first is gross margin. Nvidia's own second-quarter guide is 75.0% non-GAAP and 74.9% GAAP, each plus or minus 50 basis points. The Investing.com preview named margin as a key watch item on the argument that compression would signal pricing pressure from custom AI silicon — though it cited 73.5% as the level to watch, which does not match the company's own 75.0% non-GAAP guide, and the preview did not explain the discrepancy. The guided figures are the ones Nvidia published.

The second is China. The Investing.com preview said H20 export approval had come through but that no orders had followed. Nvidia's guidance already assumes nothing from Data Center compute in China, so the asymmetry runs one way: any China contribution would be upside to the guided number, and its absence is already in the guide. Whether management changes that assumption for the October quarter would be a more consequential disclosure than the quarter's revenue line.

There is a capital-returns dimension too. Alongside first-quarter results, Nvidia raised its quarterly cash dividend from $0.01 to $0.25 per share and authorised an additional $80 billion of share repurchases. For a company generating cash at this rate, buyback pace is one of the few levers that speaks to management's own view of the durability of demand, and it is worth reading in the same breath as the guidance.

The report lands into a tape that has been choppy rather than euphoric. AP's Friday market wrap put the Nasdaq composite at 26,180.45, up 113.29 points or 0.43%, the S&P 500 at 7,674.37 (+0.43%), the Dow at 53,277.01 (+0.98%) and the Russell 2000 at 3,017.87 (+0.85%) — with all four indexes still finishing the week lower. The Investing.com preview also noted Bank of America's Bull & Bear Indicator at 9.5, a level it described as deep in extreme sell territory. That is a positioning observation attributed to that preview, not a forecast.

For the smaller names across memory, optics, power and substrates that trade on Nvidia's disclosures, the useful signal on Wednesday is not the headline revenue figure. It is the networking line, the margin line and the sequential step embedded in the October guide — three numbers that between them describe how much of the buildout is still ahead. None of them exist yet.