NVIDIA FY2026 Q2 Revenue Hits $96.2B, Guidance Exceeds $100B for the First Time
Nashnova编辑部
Nvidia posted $96.2 billion in quarterly revenue — up 106% year-on-year — and guided above $100 billion for the first time; but free cash flow halved to $21 billion as receivables surged, exposing lengthening payment cycles from its largest customers.
How much did Nvidia actually earn?
Revenue hit $96.2 billion, up 106% year-on-year, beating the S&P Capital IQ consensus of $92.1 billion. Net income rose 126% to $59.7 billion.
Adjusted EPS came in at $2.22, above the Street's $2.09 estimate. This means → Nvidia's profitability is still accelerating, not just riding top-line scale.
Gross margin held at 75%, in line with expectations. Operating margin reached 66.5%, above last year's 64.5%. Operating expenses rose 55% to $8.4 billion — spending more, but earning far more.
Why does the data-center segment account for over 90%?
Data-center revenue reached $89 billion for the quarter, up 117% year-on-year, making up 93% of total revenue.
Hyperscaler revenue — from the likes of Amazon AWS and Microsoft Azure — grew 102%, but the rest of the customer base grew even faster at 138%.
This means → Nvidia's revenue mix is diversifying away from a handful of giants. In plain terms = the old fear that "one hyperscaler cutting orders could tank the business" is weakening — smaller buyers are now growing faster.
First guidance above $100 billion — where does the confidence come from?
Nvidia guided next quarter (August–October) revenue to roughly $108 billion, plus or minus 2% — the company's first-ever quarter above $100 billion.
That tops analysts' prior consensus of about $104.9 billion, implying year-on-year growth of roughly 89%.
This means → management has unusually high demand visibility one quarter out. Amazon's announcement that it will deploy an additional 2 million GPUs — which briefly pushed Nvidia shares up about 3% after hours — reinforces that read.
Why did free cash flow suddenly halve?
Free cash flow dropped more than 50% quarter-on-quarter to $21 billion, driven by a roughly 50% jump in accounts receivable.
Nvidia attributed this to "extended payment terms in large, multi-quarter agreements with certain investment-grade customers." In plain terms = the chips shipped, but the cash hasn't come back — the bigger the buyer, the longer the payment cycle.
This reflects a structural shift: when buyers are major cloud providers, Meta, and SpaceX, Nvidia's pricing power shows up in the unit price, but the cash-collection tempo is set by the buyer. Whether this number recovers in coming quarters is the metric the market will watch most closely.
Why is Nvidia signing leases and making equity bets?
The company disclosed for the first time that it has signed $20 billion in large, long-term data-center leases, which it plans to assign to customers.
Its equity-investment portfolio is now worth $99 billion, with a further $25 billion committed. Nvidia has already pledged $30 billion in direct investment in OpenAI and $108 billion in credit support, backing OpenAI's plan to lease a major Ohio data center starting in 2028.
This means → Nvidia is extending from "selling chips" to "locking in the ecosystem" — leases bind customers, equity stakes bind developers. This reflects an AI-infrastructure race that has entered the phase of "who ties customers into their stack first."
After a year of gains, is the stock expensive?
Nvidia shares slipped in after-hours trading, extending a seven-session losing streak, but remain up roughly 11–12% year-to-date.
The stock now trades at about 20.8× earnings, down from 26.7× three months ago and 36× a year ago. In plain terms = the share price is near highs, but because profits have grown faster, the valuation multiple has actually compressed.
The unresolved question: with free cash flow plunging and receivables piling up, the market will keep asking — is this growth backed by real cash, or by a "ship now, collect later" paper boom?
市场有风险，内容仅供研究参考，不构成投资建议。