NVIDIA's Next-Generation Rubin Chip Lineup Unveiled, Huang Predicts AI Chips Could Generate $1 Trillion in Revenue by 2027

K-LinePoet · 2026-03-24T03:42:36+00:00

Nvidia CEO Jensen Huang unveiled new hardware at GTC2026, including the Vera processor, predicting that AI chip sales will generate at least $1 trillion in revenue. He emphasized surging computing demands and showcased the new Vera CPU and liquid-cooled server architecture, demonstrating support for building "AI Factory" strategy.

K-LinePoet

2026-03-24 03:42:36

Abstract generation in progress

IT House, March 17 — NVIDIA CEO Jensen Huang delivered a series of major announcements during his keynote at the 2026 GTC Technology Conference.

He not only announced several new hardware products, including the all-new Vera processors, but also significantly raised the company’s AI chip sales forecast. He predicts that Blackwell and Rubin series products will generate at least $1 trillion in revenue (IT House note: approximately 6.91 trillion RMB at current exchange rates), further solidifying NVIDIA’s leadership in AI computing.

During his keynote at GTC 2026, Huang stated that in the coming months, sales of Blackwell and Rubin chips will become a major revenue source for the company. “I see AI chip sales reaching at least one trillion dollars by 2027,” he said. This figure drew gasps from the audience. Previously, NVIDIA had forecasted that by the end of 2026, its data center equipment would generate $500 billion in sales (approximately 3.46 trillion RMB at current exchange rates).

Huang attributed this revenue surge to explosive growth in computing demand. “I believe that over the past two years, computing demand has increased a millionfold,” he said. “This is felt by all of us, and by every startup.”

To sustain this growth momentum, Huang showcased or announced a series of major products, including at least seven new Vera Rubin chips. The most notable is a new Vera CPU, scheduled for release in the second half of 2026. NVIDIA claims this CPU is specifically designed for AI agents, offering double the energy efficiency, 50% faster speeds, and the industry’s highest single-thread performance and per-core bandwidth.

Additionally, NVIDIA unveiled a new server rack system integrating 256 liquid-cooled Vera CPUs. This setup can support over 22,500 concurrent CPU environments, each capable of running at full speed independently. This is a key part of NVIDIA’s strategy to build an “AI factory,” aiming to provide computing power for a range of applications from quantum computing to robotics.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.