At Nvidia’s annual GTC developer conference in San Jose, CEO Jensen Huang delivered a striking forecast that underscores the insatiable demand for AI computing infrastructure. The company now projects $1 trillion in cumulative orders for its Blackwell and Vera Rubin systems through 2027, up from approximately $500 billion in demand projected just last year for the 2025-2026 period.
Vera Rubin Enters Full Production
The Vera Rubin platform, Nvidia’s next-generation GPU architecture, is now in full production. The system comprises seven chips that have all reached mass production status. This marks a significant milestone as Nvidia transitions from the recently launched Blackwell architecture to its successor in record time.
The timing is notable: Blackwell chips are still ramping up availability, yet Vera Rubin is already in full production—a reflection of the unprecedented pressure on AI compute supply.
10x Inference Cost Reduction
A key highlight from Huang’s keynote was the performance profile of Vera Rubin. The new platform promises up to 10x reduction in inference token cost compared to Blackwell, addressing one of the most pressing economics challenges in AI deployment. As inference workloads dominate AI usage patterns, this efficiency gain could significantly accelerate enterprise adoption.
Groq LPU Integration
During the conference, Nvidia also announced a partnership with Groq to integrate its Language Processing Units (LPUs) into the Nvidia ecosystem. This collaboration adds another option for developers seeking specialized inference hardware beyond traditional GPUs.
Market Sentiment Reset
The announcements appear to have reset investor sentiment after earlier concerns about Blackwell demand. Nvidia’s stock rose during the conference as the $1 trillion cumulative revenue projection through 2027 signaled sustained growth horizons that far exceed typical chip industry cycles.
With over 30,000 developers attending GTC 2026 in person, the event reinforced Nvidia’s position as the central infrastructure provider for the AI industry—at a moment when demand shows no signs of abating.