Prefer to watch? Here’s the summary:
Wednesday, November 18, 2026
Wall Street has a peculiar habit of treating extraordinary success as an invitation to demand more extraordinary success. Nvidia has obliged. Third-quarter revenue reached $115.8 billion, more than $5 billion above the top of its own forecast. Margins, operating income and earnings exceeded our strong lines. For a company already valued on expectations of technological dominance, the performance was remarkable. Unfortunately, success has a way of creating obligations that enthusiasm cannot discharge.
The quarter also arrived under circumstances no spreadsheet could anticipate. Six days after Jensen Huang resigned, interim chief executive Colette Kress presented the results of the company he built. The machinery of artificial intelligence continued humming, customers continued buying and profits continued accumulating. Yet investors now confront a question more complicated than whether Nvidia can sell another billion dollars of chips: Can the institution outlast the personality that defined it? The quarter answered questions about the business and said little about its future leadership.
The Scorecard: Five Bulls and Two Blanks
We published eight lines on November 12, before Nvidia reported November 17. These are not attempts to predict precise results. They establish what constitutes strong, middling or weak performance before the numbers become known. We compare outcomes with company disclosures rather than analyst estimates. A result at or beyond the strong line earns a Bull grade; at or beyond the weak line, Bear; everything between is Neutral. We do not change lines after earnings, and undisclosed measures remain ungraded. The purpose of scorecards is not to prove that we can see the future. It is to prevent ourselves from rewriting the past.
Revenue: Strong (Bull) line ≥ $110.16B; Weak (Bear) line ≤ $105.84B; Result $115.8B; Grade: Bull
Data Center revenue: Strong (Bull) line ≥ $102B; Weak (Bear) line ≤ $98B; Result $107.6B; Grade: Bull
GAAP/non-GAAP gross margin: Strong (Bull) line ≥ 74.5%; Weak (Bear) line ≤ 73.5%; Result 74.8% / 74.9%; Grade: Bull
GAAP operating income: Strong (Bull) line ≥ $72.87B; Weak (Bear) line ≤ $68.59B; Result $77.52B; Grade: Bull
Non-GAAP diluted EPS: Strong (Bull) line ≥ $2.54; Weak (Bear) line ≤ $2.36; Result $2.68; Grade: Bull
Rubin share of Data Center: Strong (Bull) line ≥ 21%; Weak (Bear) line ≤ 19%; Result Undisclosed; Grade: Ungraded
Q4 gross margin guidance: Strong (Bull) line ≥ 72%; Weak (Bear) line ≤ 71%; Result 71%–72%; Grade: Neutral
Groq 3 LPX deployment: Strong (Bull) line On schedule, expanded; Weak (Bear) line No Q3 launch; Result Undisclosed; Grade: Ungraded
Result: 5 Bull, 1 Neutral, 0 Bear and 2 Ungraded.
The distinction matters. A scorecard is useful only when its author resists the temptation to improve yesterday’s predictions with today’s information. Wall Street has enough historians who specialize in predicting the past.
A Business That Makes Its Forecasts Look Timid
Revenue increased 103% from a year earlier and 20% sequentially. Data Center sales reached $107.6 billion, exceeding our estimated strong line by $5.6 billion. GAAP operating income of $77.52 billion cleared its line by $4.65 billion while non-GAAP earnings reached $2.68 per share against our $2.54 Bull line.
Gross margins remained impressive despite the costs of introducing Vera Rubin. GAAP margin was 74.8% and non-GAAP margin was 74.9%, both above our 74.5% line. GAAP earnings reached $2.70 per share, helped by $1.8 billion in net gains on equity securities. Our scorecard uses non-GAAP earnings because gains on an investment portfolio are not the same as selling more processors.
These results establish something important: Huang’s departure has not interrupted Nvidia’s immediate commercial momentum. They do not establish how long that momentum can endure.
Two Questions Without Numbers
Nvidia says its Vera Rubin NVL72 system has begun shipping and describes the introduction as its fastest product ramp. “In its first quarter of production shipments, the Rubin platform generated nearly $20 billion in revenue, making it the fastest product ramp in NVIDIA’s history,” Kress said on the call.
The company also offered no reported status for Groq 3 LPX, the inference platform Nebius was expected to adopt first. Investors may draw conclusions from silence. Our scorecard does not.
The Cost of Tomorrow’s Growth
Fourth-quarter revenue guidance is $130 billion, plus or minus 3%, including an extra reporting week and assuming no Data Center compute revenue from China. Gross margin guidance of 71% to 72% earns Neutral under our published lines. Rising memory costs, which Nvidia warned about in August, remain a pressure on profitability.
More revealing are the obligations accumulating behind the income statement. Supply commitments increased from $279 billion to $390 billion in one quarter, principally for multiyear memory purchases. Nvidia also guarantees up to $118 billion in partner obligations, including as much as $105 billion supporting SB Energy’s Ohio campus, which will be leased to OpenAI.
Accounts receivable reached $81 billion while days sales outstanding increased from 60 to 64. Nvidia attributes part of the change to extended payment terms on large agreements. Quarterly operating cash flow of $48.3 billion provides substantial financial capacity, but commitments of this magnitude require more than cash. They require judgment about customers, technology and demand years into the future.
The company is increasingly financing the ecosystem that purchases its products. This may prove visionary. It also makes the distinction between customers’ demand and customers’ ability to finance that demand increasingly important.
The Question Still Unanswered
Nvidia’s disclosures now acknowledge its leadership transition, the search for a permanent chief executive and an independent board committee investigation. None can be resolved by another record quarter.
The forthcoming 10-Q may provide additional detail about Rubin and LPX. If it does, we will grade the outstanding lines without changing them. Our fourth-quarter scorecard will likewise be published before the results arrive.
Nvidia has demonstrated that the machine Huang built continues to operate brilliantly. What remains uncertain is who will steer it through the financial commitments, technological transitions and competitive pressures that its extraordinary success has created.
The engines are running. The destination is ambitious. The question is who now holds the map.
Sources: Nvidia third-quarter fiscal 2027 earnings release and CFO commentary, November 17, 2026; third-quarter earnings call, November 17, 2026; Form 8-K, November 12, 2026; second-quarter earnings call, August 26, 2026; Portfolio Lens Research, November 12, 2026.
Lines first. Grades later. Misses included.
Not investment advice. Research for informational purposes only. Investing involves risk, including loss of principal. Past performance does not indicate future results.

