NVIDIA has confirmed that its next generation Vera Rubin AI platform is scheduled to begin initial shipments in the third quarter of 2026, with volume ramp expected in the fourth quarter. The company also expects shipments to grow further through the first half of 2027 as demand for AI infrastructure continues to rise.
This is related to the earlier Vera CPU and Vera Rubin rack cost stories, but it is a different update. The earlier reports focused on standalone Vera CPU revenue and the rising bill of materials for Vera Rubin racks. This report focuses on NVIDIA’s launch timeline for Vera Rubin and the continued demand for Blackwell systems.
NVIDIA reported record revenue of $81.6 billion for Q1 FY2027, up 85 percent year over year. Data Center remained the main driver with $75.24 billion in revenue, while Edge Computing reached $6.36 billion.
Blackwell remains in heavy demand while NVIDIA prepares the next AI platform
NVIDIA said demand for Blackwell GB300 and NVL72 systems remains strong among frontier model builders and hyperscalers. The company said hundreds of thousands of Blackwell GPUs have already been deployed globally, and partner data centers above 10 megawatts have nearly doubled in one year to more than 80 sites.
The demand is also pushing up prices for older NVIDIA AI GPUs. H100 rental pricing has risen around 20 percent year to date, while A100 cloud pricing is up nearly 15 percent. This shows that AI demand is still spreading across both current and previous generation hardware.
| Area | Details |
|---|---|
| Q1 FY2027 revenue | $81.6 billion |
| Data Center revenue | $75.24 billion |
| Edge Computing revenue | $6.36 billion |
| Vera Rubin first shipments | Q3 2026 |
| Vera Rubin volume ramp | Q4 2026 |
| Wider shipment growth | First half of 2027 |
| Blackwell deployment | Hundreds of thousands of GPUs |
| Large partner data centers | More than 80 sites above 10 megawatts |
NVIDIA CEO Jensen Huang said Vera Rubin is already off to a strong start and is expected to be more successful than Grace Blackwell. He also said frontier model companies are expected to adopt Vera Rubin from the beginning, which suggests NVIDIA already has demand lined up before the platform reaches volume production.
The company is also mass producing Vera CPUs for rack deployments. Early shipments have already gone to customers such as OpenAI, SpaceX, Oracle, and Anthropic.
NVIDIA’s current position is strong because Blackwell is still seeing heavy demand while Vera Rubin is moving toward launch. The key question now is how quickly NVIDIA can scale supply once Vera Rubin enters its ramp phase in late 2026.
Discussion (0)
Be the first to comment.