Sources of revenue in Nvidia's Q2 fiscal 2027, the quarter ended July 26, 2026, and what the mix says about AI infrastructure demand.

Research report
Sources of revenue in Q2 fiscal 2027, the quarter ended July 26, 2026.
All segment and platform figures are from the company's Q2 FY27 press release and CFO commentary filed with the SEC on August 26, 2026. Earnings-call figures are as stated by management on the same day's call and are not separately audited.
Inside
Revenue of $96.2 billion was up 18% sequentially and 106% from a year ago, the fourth consecutive quarter of accelerating year-on-year growth. It beat the analyst consensus of $92.1 billion and Nvidia's own guidance midpoint of $91 billion, the fourteenth straight quarter above the company's outlook. First-half fiscal 2027 revenue reached $177.8 billion, already 82% of what the company earned in all of fiscal 2026.
Nvidia reports two segments. Compute & Networking carried $88.3 billion of the quarter, or 92% of the total, and grew 114% year on year. Graphics contributed $7.9 billion and grew 46%.
| Segment | Q2 FY27 | Y/Y | Share of revenue |
|---|---|---|---|
| Compute & Networking | $88.3B | +114% | 92% of total |
| Graphics | $7.9B | +46% | 8% of total |
The more useful cut is by market platform, which the CFO commentary splits into Data Center and Edge Computing, with Data Center further divided into two customer groups: Hyperscale, and AI Clouds, Industrial & Enterprise (ACIE). During the quarter Nvidia reclassified one company from ACIE to Hyperscale after a change in its business model, and recast prior periods to match.
| Platform | Q2 FY27 | Q/Q | Y/Y |
|---|---|---|---|
| Data Center | $89.0B | +18% | +117% |
| Hyperscale | $48.7B | +13% | +102% |
| AI Clouds, Industrial & Enterprise | $40.3B | +25% | +138% |
| Edge Computing | $7.2B | +13% | +27% |
Hyperscale, $48.7 billion. The largest cloud operators more than doubled their spend from a year ago and grew it 13% sequentially on the strength of Blackwell Ultra. Hyperscale is now 55% of Data Center revenue, down from about 57% last quarter as the other group grows faster.
ACIE, $40.3 billion. AI clouds (the NeoClouds), industrial and enterprise customers grew 138% year on year and 25% sequentially, driven by AI natives, enterprises and sovereign customers, plus hyperscalers renting capacity from AI clouds. Within it, sovereign AI revenue grew 35% sequentially and more than tripled from a year ago. On the call, Huang put ACIE at roughly half the business and growing at an annual rate of 100%.
Edge Computing, $7.2 billion. Up 27% from a year ago on strong Blackwell workstation sales, partly offset by slower consumer PC sales as elevated memory prices pushed up system costs.
Two smaller threads matter for the mix. Networking revenue set a record, up 18% sequentially, with Spectrum-X Ethernet revenue growing 2.6 times year on year; trailing-twelve-month Grace CPU revenue passed $5 billion. And China has effectively left the picture: Hopper shipments to China were under 1% of Data Center revenue, and the forward outlook assumes no China data center compute at all.
Customer concentration is easing in one direction only. The non-hyperscale half of the data center business is growing twice as fast as the hyperscale half, and sovereign AI tripled in a year.
Six quarters ago Nvidia reported $44.1 billion. The quarter just closed was $96.2 billion, and guidance puts the current quarter at $108 billion. Growth has not just continued at scale, it has accelerated for four quarters running.
Gross margin held at 75.0% on both GAAP and non-GAAP bases, up 2.6 points from a year ago on richer Blackwell Ultra mix. Operating expenses of $8.4 billion grew 55% year on year, far slower than revenue, so GAAP operating income more than doubled to $63.7 billion. Net income of $59.7 billion included $7.8 billion of gains on equity securities.
Free cash flow was $21.3 billion, and Nvidia returned a record $26 billion to shareholders in the quarter, about $20 billion in buybacks and $6 billion in dividends. The balance sheet is being put to work ahead of the Vera Rubin ramp: inventory rose to $31.6 billion, supply commitments jumped from $119 billion to $279 billion (mostly memory procurement), and days sales outstanding stretched to 60 on extended payment terms for large multi-quarter deals.
Guidance for Q3 fiscal 2027 is $108 billion, plus or minus 2%, with no China data center compute assumed. Vera Rubin, now in full production, is expected to be Nvidia's fastest ramp yet and to account for about 20% of data center revenue in Q3. On the call, CFO Colette Kress said the company expects revenue to grow approximately 70% in fiscal 2028, and called that a supply-constrained outlook.
| Measure | Q3 FY27 outlook | Note |
|---|---|---|
| Revenue | $108.0B, plus or minus 2% | No China DC compute assumed |
| Gross margin | 74.0%, plus or minus 50 bps | Q4 expected to bottom at 71-72% |
| Operating expenses | ~$9.2B GAAP, ~$9.0B non-GAAP | FY27 tax rate 16-18% |
| Fiscal 2028 revenue | Approximately +70% | Supply-constrained, per CFO |
The margin path is the trade-off. Q3 gross margin is guided to 74.0%, Q4 is expected to bottom in the 71% to 72% range as memory costs bite, and fiscal 2028 is expected to settle at 72% to 73%. Memory pricing was the one pressure management flagged repeatedly, describing "extreme pricing conditions" that have run past its earlier expectations.
All segment and platform figures are from the company's Q2 FY27 press release and CFO commentary filed with the SEC on August 26, 2026. Earnings-call figures (sovereign AI, networking, Grace CPU, fiscal 2028 guidance) are as stated by management on the same day's call and are not separately audited.
Build grounded agents
See how lowtouch.ai turns enterprise rules, policies, and semantic context into governed agents running inside your appliance.
About the Author

Pradeep Chandran
Lead - Agentic AI & DevOps
Pradeep Chandran is a seasoned technology leader and a key contributor at lowtouch.ai, a platform dedicated to empowering enterprises with no-code AI solutions. With a strong background in software engineering, cloud architecture, and AI-driven automation, he is committed to helping businesses streamline operations and achieve scalability through innovative technology. At lowtouch.ai, Pradeep focuses on designing and implementing intelligent agents that automate workflows, enhance operational efficiency, and ensure data privacy. His expertise lies in bridging the gap between complex IT systems and user-friendly solutions, enabling organizations to adopt AI seamlessly. Passionate about driving digital transformation, Pradeep is dedicated to creating tools that are intuitive, secure, and tailored to meet the unique needs of enterprises.