What To Expect When NVIDIA (NVDA) Releases Q3 Earnings

NVIDIA (Nasdaq: NVDA | NVDA Price Prediction) reports fiscal third quarter 2026 results after the close, and expectations could hardly be higher. The company is coming off another record quarter with $46.7 billion in Q2 revenue, driven by surging demand for Blackwell, strong sequential growth in networking, and broad adoption across CSPs, enterprises and sovereign AI programs. Jensen Huang and Colette Kress emphasized that the global AI build-out is still in its early innings, with industry CapEx now running at an astonishing $600 billion per year. As the AI race intensifies, NVIDIA remains the core infrastructure provider to hyperscalers, AI-native startups and industrial users.

Analysts expect another strong quarter, with both EPS and revenue projecting substantial double-digit growth. But the bigger story is NVIDIA’s positioning heading into 2026, as Blackwell continues to scale and the company prepares for Rubin’s volume launch next year. With AI factories expanding from tens of megawatts toward hundreds of megawatts, NVIDIA’s platform-level strategy across compute, networking and software keeps it central to global AI deployment.

What to Expect When NVIDIA Reports

NVIDIA is expected to post year-over-year revenue growth of 57 percent and EPS growth of 55 percent, underscoring the scale of AI infrastructure demand. Full-year estimates also continue to push higher following persistent guidance raises and sequential momentum.

Key Areas to Watch

1. Blackwell ramp and GB300 production cadence
Management noted that Blackwell reached record levels in Q2, with the GB300 transition described as “seamless” and production now running at roughly 1,000 racks per week . Investors will want updated commentary on how quickly GB300 capacity is scaling and whether new CSP deployments are accelerating into Q4 and early 2026.

2. Networking growth across NVLink, InfiniBand and Spectrum-X
Networking revenue reached a record $7.3 billion last quarter, with NVLink, Spectrum-X and InfiniBand all posting exceptional gains. NVIDIA emphasized that networking efficiency can effectively determine AI factory economics, particularly as single facilities approach gigawatt scale. Any signal of continued triple-digit growth or constraints will heavily influence forward expectations.

3. Impact of China licensing and H20 uncertainty
NVIDIA’s Q3 outlook excluded all H20 shipments to China due to ongoing geopolitical review. Management indicated the quarter could see $2–5 billion in H20 revenue if licensing proceeds, but visibility remains limited . Investors will listen closely for any update on approvals, potential Blackwell eligibility in China, and the long-term revenue profile of the region.

4. AI infrastructure demand and the $3–4 trillion build-out
Jensen Huang reiterated that the world is still “at the beginning of an industrial revolution,” projecting $3–4 trillion in AI infrastructure spend by decade-end. With hyperscaler CapEx already doubling to $600 billion annually, comments on 2026 demand signals, power constraints, and multi-gigawatt AI factory planning will be key.

5. Rubin platform progress and annual cadence confirmation
Rubin chips—including Vera CPU, Rubin GPU, and CX9 SuperNIC, are already in fab and on schedule for volume next year . NVIDIA emphasized its shift to a consistent annual product cadence. Investors will look for additional clarity around performance uplift, supply-chain readiness, and customer ramp expectations.