NVIDIA second quarter fiscal 2027 earnings explained: segments, EPS and valuation multiples

NVIDIA reported results for the second quarter of its 2027 fiscal year on August 26, 2026. NVIDIA's fiscal year ends in late January, so this quarter covers the three months ended July 26, 2026. As of November 1, 2026 NVIDIA had not yet reported its third quarter of fiscal 2027, so these are its latest published results. Revenue rose 17.9% from the quarter before and operating income rose 19.0%, while GAAP net income rose only 2.3% (Dalsna Finance calculation). This article walks through what NVIDIA does, how its businesses performed and what the numbers say about its valuation. Every figure is sourced and dated, and every number worked out by us is labelled as a Dalsna Finance calculation. ^[1][2]

The share price used further down is the close on October 30, 2026, more than two months after these results. It reflects everything the market learned in between, including the deals NVIDIA announced after the quarter ended, which are covered below. ^[3][4][5]

NVIDIA in brief

NVIDIA Corporation, based in Santa Clara, California, designs computing platforms. Its quarterly report says it started out focused on PC graphics and used its GPU (graphics processing unit) designs to build platforms for scientific computing, AI, data science, autonomous vehicles, robotics and digital twins. It now describes itself as a data center scale AI infrastructure company. Its shares trade on the Nasdaq under the ticker NVDA, and StockAnalysis lists their first trading day as January 22, 1999. ^[2][1][3]

In the release, founder and chief executive Jensen Huang said "The AI infrastructure buildout is at full steam." He also said demand is accelerating and that Vera Rubin, NVIDIA's newest platform, is now in full production (an attributed paraphrase). ^[1]

Two segments and two market platforms

NVIDIA reports two segments. Compute & Networking covers its Data Center accelerated computing and networking platforms, AI solutions and software, plus automotive platforms and autonomous and electric vehicle solutions. Graphics covers GeForce GPUs for gaming and PCs and Quadro/NVIDIA RTX GPUs for workstation graphics in businesses. In the second quarter Compute & Networking had operating income of $62,696 million and Graphics $3,899 million, as stated in the quarterly report. Together that is more than the company's $63,734 million of operating income because some expenses, such as stock based pay and corporate infrastructure and support costs, are not allocated to either segment. ^[2][6]

NVIDIA also splits revenue into two market platforms: Data Center and Edge Computing. Data Center is divided into Hyperscale and a group NVIDIA calls ACIE, short for its own name AI Clouds, Industrial, & Enterprise. In the second quarter NVIDIA moved one company from ACIE to Hyperscale because that company's business model changed, and it recast the earlier figures to match. Products named in the release include the Vera Rubin platform, the Vera CPU, the Groq 3 LPX inference accelerator, RTX Spark (a superchip for Windows PCs that is part of a partnership with Microsoft) and the DRIVE Hyperion platform for robotaxis. ^[6][1]

Where the revenue comes from

Data Center revenue was $89,023 million, 92.5% of the total (Dalsna Finance calculation). NVIDIA said it grew both from a year earlier and from the first quarter, driven by the ramp of its Blackwell Ultra infrastructure. Within it, Hyperscale grew 13% from the first quarter, which NVIDIA put down to Blackwell Ultra, and ACIE grew 25%, which it credited to demand from AI native companies, businesses and governments, plus hyperscalers renting capacity on AI clouds. NVIDIA said Data Center Hopper products shipped to China were less than 1% of Data Center revenue in the quarter. ^[6][1]

Edge Computing revenue was $7,198 million, up 13% from the first quarter and 27% from a year earlier. NVIDIA said strong sales of Blackwell workstations drove both increases, partly offset by slower sales of consumer PCs, which it said were held back by higher memory and system prices. ^[6]

The numbers, quarter over quarter

Revenue was $96,221 million in the second quarter of fiscal 2027, against $81,615 million in the first. Gross profit, revenue minus the direct cost of what was sold, rose 18.0% to $72,142 million. Operating income, which also takes off research, sales and administrative costs, rose 19.0% to $63,734 million. Net income under GAAP, the standard US accounting rules, rose 2.3% to $59,688 million (Dalsna Finance calculation for the changes). Against the second quarter of fiscal 2026, revenue was up 106% and net income up 126%, as reported by NVIDIA. ^[1]

Net income grew much more slowly than operating income because of other income, which sits below the operating line. It was $16,367 million in the first quarter, of which $15,936 million was net gains on equity securities, the shares NVIDIA holds in other companies. In the second quarter other income was $7,773 million, with $7,771 million of such gains. NVIDIA's effective tax rate was 16.5%, which it said was higher than a year earlier mainly because revenue was higher. ^[1][6]

NVIDIA also reports adjusted (non GAAP) figures. They leave out gains and losses on equity securities, acquisition related and other costs, some other income and expense and the tax effect of those items. From the first quarter of fiscal 2027 they no longer leave out stock based pay, and NVIDIA updated its earlier non GAAP figures to match. Gross margin was about flat from the first quarter, which NVIDIA said was because its Blackwell architecture still makes up the great majority of revenue. GAAP operating expenses rose 10.3% (Dalsna Finance calculation) and adjusted operating expenses rose 11%, as reported, which NVIDIA said came from higher compute infrastructure costs and higher compensation and benefits. ^[1][6]

Earnings per share

Earnings per share, or EPS, is profit divided by the number of shares. The diluted figure also counts shares that could be created from stock awards. NVIDIA reported diluted EPS of $2.46 for the second quarter of fiscal 2027 and $2.39 for the first. Dividing net income by the diluted share count gives the same figures. Adjusted diluted EPS, as reported, was $2.22 and $1.87. The gap between the two measures was narrower in the second quarter, when net gains on equity securities were smaller. ^[1]

What the market values NVIDIA at

Market capitalization, or market cap, is what the stock market values the whole company at: the share price times the number of shares. NVIDIA's shares closed at $204.64 on Friday, October 30, 2026, as reported by StockAnalysis. The cover of its second quarter report to the SEC lists 24.1 billion shares outstanding as of August 21, 2026. That gives a market cap of about $4.93 trillion (Dalsna Finance calculation). The cover rounds the share count to the nearest 0.1 billion, so the market cap could be off by about $10 billion either way, which moves the multiples below by less than 0.1. ^[3][2]

Six valuation multiples

The price to earnings ratio, or P/E, is market cap divided by a year of earnings. It shows how many dollars the market is paying for each dollar of yearly profit. Annualised means one quarter multiplied by four. On that basis NVIDIA has a P/E of 20.7 (Dalsna Finance calculation: $4,931.8 billion divided by $59,688 million times 4, which is $238,752 million). On adjusted net income times 4, $215,816 million, the adjusted P/E is 22.9. ^[1][3][2]

A forward P/E uses an estimate of future earnings. NVIDIA's outlook for the third quarter covers revenue, gross margin, operating expenses and the tax rate, but not earnings, and we do not turn it into an earnings figure ourselves. We did not find a full year earnings estimate published after the August 26, 2026 results that we could check against its original source. So no forward P/E or adjusted forward P/E is calculated. ^[1]

A trailing P/E, sometimes called a historical P/E, uses the last four reported quarters. GAAP net income over the four quarters to July 26, 2026 added up to $192,879 million, which gives a trailing P/E of 25.6 (Dalsna Finance calculation). There is no adjusted trailing P/E. NVIDIA has published its new adjusted figures, which include stock based pay, for the last three of those quarters but not for the third quarter of fiscal 2026, and adding figures made on two different bases would not give a fair total. ^{[1][7][8][3][2]}

The last four quarters

Each of the last four quarters is reported in its own earnings release. GAAP net income rose every quarter, from $31,910 million in the third quarter of fiscal 2026, the three months ended October 26, 2025, to $59,688 million in the second quarter of fiscal 2027. The adjusted column shows NVIDIA's figures on its current basis, which includes stock based pay. For the third quarter of fiscal 2026 NVIDIA has only published an adjusted figure on its old basis, $31,767 million, which left stock based pay out, so it is not shown in the column or the chart. ^[8][7][1]

Three base cases: the second quarter held flat for a year

Suppose revenue stayed at the second quarter's level for a full year, with no growth and no decline: $96,221 million times 4, or $384,884 million. The first base case keeps the 62.0% profit margin, which gives earnings of $238,752 million, or $9.91 per share on the cover share count. The second keeps the 56.1% adjusted profit margin, which gives $215,816 million, or $8.96 per share. Each case is simply the latest quarter annualised, so at its own annualised P/E it returns the October 30, 2026 close of $204.64 by construction. The other cells only show how the price moves when the same earnings are given a different multiple. ^[1][3][2]

Second quarter net income was lifted by items below the operating line, mainly the $7,771 million of net gains on equity securities, which NVIDIA's CFO commentary rounds to $7.8 billion. So the third base case uses the 66.2% operating margin, which leaves out interest, taxes and other income altogether: $254,936 million a year, or $10.58 per share. Operating income is counted before tax, so this is not net income. NVIDIA said its H20 charges and releases were insignificant in the second quarter, and we found no other one time item inside operating income that it put a figure on, so nothing else is taken out. ^[1][6]

Cash, debt and commitments

Cash from operations was $24,077 million in the second quarter, down from $50,344 million in the first, and free cash flow, cash from operations minus spending on property, equipment and intangible assets and the related principal payments, was $21,341 million, as reported by NVIDIA. NVIDIA said the drop came from bigger working capital changes and higher cash taxes. Accounts receivable rose to $63,059 million, or 60 days of sales from 45, which it said reflects longer payment terms on large agreements with certain investment grade customers. Inventory rose to $31,575 million as NVIDIA prepared to introduce Vera Rubin in the third quarter. ^[1][6]

NVIDIA said it returned about $26.0 billion to shareholders in the quarter through share repurchases and dividends. Its cash flow statement shows $19,732 million paid for repurchases of common stock and $6,047 million of dividends paid, against $244 million of dividends a year earlier. It issued $25.0 billion of senior unsecured notes for general corporate purposes, and long term debt was $32,366 million at July 26, 2026. Its commitments to suppliers rose to $279 billion from $119 billion, mainly for memory, as stated in the CFO commentary and the quarterly report. ^[1][6][2]

After the quarter ended

On August 17, 2026 NVIDIA announced a partnership with SB Energy for the PORTS Technology Campus in Pike County, Ohio, where an affiliate of OpenAI will be the tenant, and signed residual value guaranties on leases for about 4.25 gigawatts of capacity there. NVIDIA would pay only if OpenAI became insolvent and defaulted on a lease or failed to make lease payments, and then broadly the shortfall between a guaranteed minimum lease value and what a new lease or a sale recovers. Its payments are capped at $105 billion in total for this initial commitment and take effect in phases as conditions are met, with the first expected in fiscal 2029 (a company expectation). NVIDIA can also choose to provide credit support for about 3.8 more gigawatts at the site. In August it also signed memoranda of understanding with large capital providers on financing platforms for AI infrastructure, and its quarterly report says these may not lead to definitive agreements. ^[4][6][2]

On September 2, 2026 NVIDIA agreed to acquire Hugging Face, whose platform and community let developers create, share and put to use open source models, datasets and applications, for a purchase price of about $11.9 billion, subject to certain adjustments, plus a retention program of up to about $1.0 billion in equity for Hugging Face employees. NVIDIA expects the deal to close in the first half of 2027, subject to customary closing conditions including required regulatory approvals (a company expectation). ^[5]

The quarterly report also describes a securities class action first filed in 2018, which alleges misleading statements about channel inventory and cryptocurrency mining demand. On March 25, 2026 the district court certified a class of investors, and NVIDIA has recorded no liability for the case, saying a loss is reasonably possible but not probable and cannot be estimated. ^[2]

What to watch

For the third quarter of fiscal 2027 NVIDIA said it expects revenue of $108.0 billion, plus or minus 2%, with gross margins of 74.0%, plus or minus 50 basis points (a company projection). It said the outlook assumes no Data Center compute revenue from China. The next report will show how revenue compared with that outlook and whether Vera Rubin began to show up in Data Center sales. ^[1][6]

Because equity securities gains move GAAP net income from quarter to quarter, readers may want to watch operating income and adjusted net income alongside it. Receivables, inventory and supplier commitments all rose sharply in the second quarter, and the next report will show whether they kept rising. ^[1][6]

How these figures were made: inputs come from NVIDIA's earnings releases for the third quarter of fiscal 2026 and the first and second quarters of fiscal 2027, its second quarter CFO commentary, its quarterly report for the period ended July 26, 2026 and two later current reports on SEC EDGAR, plus the dated closing price named above. Calculations use unrounded inputs and are rounded only at the end, with multiples to one decimal and per share figures and prices to the cent. ^{[1][6][2][7][8][4][5][3]}