Nvidia CEO Jensen Huang Announces $1 Trillion in Blackwell and Vera Rubin Orders, Unveils

Nvidia’s Blackwell and Vera Rubin chips have already begun reshaping data‑center workloads, and Huang said the pipeline of contracts now tops $1 trillion through 2027. The figure, disclosed at GTC 2026, aggregates commitments from hyperscale cloud providers, enterprise AI teams and government research labs, according to the CNBC report on the event. Huang emphasized that the orders are “firm” and span multiple generations of the Blackwell family, including the flagship Blackwell‑X and the upcoming Vera Rubin accelerator designed for scientific simulations. The company expects to ship more than 1 billion GPU‑equivalent units by the end of 2027, a volume that would dwarf the combined shipments of its rivals, the report adds.

The order surge coincides with Nvidia’s rollout of DLSS 5, the next‑generation AI‑enhanced graphics engine that Huang called “the GPT moment for graphics.” DLSS 5, announced at the same conference, introduces a real‑time neural rendering model that injects photoreal lighting and material detail into every frame, according to the Nvidianews release. The technology will be supported by a roster of major publishers—including Bethesda, Capcom, Ubisoft and Warner Bros. Games built on DLSS 5 are expected to run on the new Blackwell GPUs, leveraging their expanded tensor cores and higher memory bandwidth to deliver Hollywood‑level visual fidelity at consumer frame rates. Huang linked the graphics breakthrough to the same AI engine that powers the Blackwell line, underscoring Nvidia’s strategy of bundling AI compute with consumer‑facing products.

Financial analysts note that the $1 trillion order book pushes Nvidia toward a market valuation that could breach the $5 trillion mark, a milestone already hinted at in recent Ars Technica coverage of the company’s growth. The article cites Huang’s dismissal of “AI bubble” concerns after Nvidia announced $500 billion in orders earlier this year, suggesting that the current $1 trillion pledge could double that figure within twelve months. While the report does not break down the geographic distribution of the contracts, it confirms that a significant share originates from Asia‑Pacific cloud operators, reflecting the region’s rapid AI adoption.

Beyond gaming, the Vera Rubin accelerator is being positioned for high‑performance scientific workloads such as climate modeling and drug discovery. According to the same CNBC piece, several national labs have placed multi‑year agreements to integrate Vera Rubin into their supercomputing clusters, citing its superior double‑precision performance and specialized interconnects. Huang highlighted that the chip’s architecture is optimized for sparse‑matrix operations, a core component of many next‑generation AI models, thereby extending Nvidia’s reach into research domains that traditionally relied on custom ASICs.

The announcements come as Nvidia prepares for its next earnings cycle, where the company will report revenue from both the AI data‑center segment and the burgeoning gaming division powered by DLSS 5. If the $1 trillion order pipeline translates into shipments on schedule, Nvidia could see annual revenue climb well beyond the $30 billion range projected by its last guidance, according to the internal projections referenced in the GTC briefing. Huang concluded by urging partners to “accelerate deployment” and promised that the Blackwell and Vera Rubin families will continue to evolve, keeping Nvidia at the forefront of both AI compute and next‑gen graphics.