NVIDIA forecasts a $1 trillion AI infrastructure buildout opportunity at its annual GTC conference.
At Nvidia's annual GPU Technology Conference (GTC) in San Jose, California, CEO Jensen Huang delivered a keynote address showcasing the company's expanding ambitions across multiple technology domains. During the two-and-a-half-hour presentation, Huang announced plans to push deeper into central processing units (CPUs)—Intel Corp's traditional territory—while introducing semiconductors incorporating technology acquired from start-up Groq Inc. The company also revealed it was developing chips for data centers in outer space. At the core of Huang's message was the assertion that computing demand has accelerated dramatically, with Nvidia positioned uniquely to meet this challenge. "I believe that computing demand has increased by 1 million times in the last two years," Huang said. "It is the feeling that we all have. It is the feeling every start-up has."
Addressing investor concerns about sustained growth, Huang unveiled a significant forecast: Nvidia's flagship AI processors would generate US$1 trillion in sales through next year, driven primarily by orders for the company's Blackwell and Rubin chips. This represents a substantial expansion of the company's previous projection that data center gear would bring US$500 billion in sales through the end of the current year. The trillion-dollar forecast effectively doubles the cumulative sales outlook and signals management's confidence in continued demand strength.
Among the product announcements, the new Groq chip emerged as a focal point, designed to enhance the responsiveness of AI systems and to be manufactured by Samsung Electronics Co, with Groq-based systems arriving in the second half. Nvidia also showcased a general-purpose CPU-based computer that could operate independently or in combination with other Nvidia systems. Huang characterized this opportunity as "for sure" a multibillion-dollar business. The company continues providing AI models and software on an open-source basis, and announced Nvidia's enterprise-focused NemoClaw system, developed with enhanced security and privacy capabilities for corporate use, following the rapid emergence of OpenClaw, an open-source agentic platform.
Looking ahead, Huang introduced the company's Feynman architecture as the next step in its technological roadmap, featuring new processing and networking technologies, including co-packaged optics. Nvidia leveraged GTC to announce partnerships with companies across multiple industries, including International Business Machines Corp, Hewlett Packard Enterprise Co, and Adobe Inc. The company also strengthened ties with Uber Technologies Inc, announcing plans for a fleet of Nvidia software-driven autonomous vehicles by 2028.
Huang emphasized Taiwan's strategic importance to Nvidia's supply chain. His presentation featured three slides indicating that among more than 60 global partners for Nvidia's Vera Rubin platform, many were Taiwanese companies including Asustek Computer Inc, Hon Hai Precision Industry Co, Quanta Cloud Technology Inc, Wistron Corp, and Wiwynn Corp. These companies will play key roles in delivering the Vera Rubin architecture to market.