Jensen Huang Predicts TSMC Must Double Production in Next Decade to Fulfill NVIDIA’s AI Demand, Underlining Future Growth in AI Industry

Taiwan Semiconductor Manufacturing Company (TSMC) is encountering significant strain on its production capabilities. NVIDIA’s CEO, Jensen Huang, has indicated that TSMC will have to more than double its production capacity to satisfy NVIDIA’s increasing demand alone.

TSMC’s Strategic Capital Investments and Facility Expansion for Unprecedented Demand

As NVIDIA propels forward in the AI sector, the company demonstrates growing confidence in the burgeoning demand for AI infrastructure over the coming years. During a conversation with regional media, Jensen announced TSMC’s plan to amplify its production capacity by over 100% in the next decade. This ambitious expansion is not only necessary for fulfilling NVIDIA’s requirements; it also signals an expectation of sustained growth in AI technologies. TSMC’s forthcoming fabrication projects and increased capital expenditures are testament to the strong demand anticipated in the semiconductor industry.

TSMC’s production capacity may grow by more than 100% in the next ten years, which is a very significant scale expansion, the largest infrastructure investment in human history, and it will have to double just to meet NVIDIA’s demand.

– NVIDIA’s Jensen Huang via UDN

In recent quarters, TSMC’s plans for factory expansion have escalated significantly, influenced by geopolitical factors. This has resulted in substantial investments to develop supply chains in regions including the EU, Japan, and the United States. For instance, TSMC is engaging in a $250 billion plan for a comprehensive facility buildout in America, which encompasses advanced packaging, semiconductor manufacturing, and research & development centers. The company’s plants in Arizona are currently transitioning to 3nm technology, with intentions to progress to A16 technology, all while adhering to the “N-2″policies.

NVIDIA and TSMC executives with Blackwell Made in America plaque on stage with American flag. — Image Credits: NVIDIA

A key insight is that NVIDIA’s initiatives, particularly through projects like Grace Blackwell and Vera Rubin, have secured a substantial portion of TSMC’s production capacity. This positioning has enabled NVIDIA to surpass Apple as TSMC’s largest client within just a few years. Additionally, TSMC’s introduction of capacity ‘prepayments’ suggests that future production lines will predominantly serve NVIDIA and other High-Performance Computing (HPC) customers, further illustrating the aggressive expansion of AI infrastructure.

NVIDIA’s dominance stems from its competitive edge in scalability, which diminishes the impact of rivals such as ASIC manufacturers and AMD. Additionally, NVIDIA’s established relationships with Taiwanese suppliers and global partners afford the company earlier access to production lines, reinforcing its leadership in the semiconductor market.

Source & Images