NVIDIA’s Explosive AI Chip Demand Pushes TSMC to Boost 3nm Production by a ‘Whopping’ 50%, Preparing for Next-Gen Rubin Lineup

Muhammad Zuhair
Image Credits: NVIDIA

NVIDIA's CEO is currently in Taiwan, and it is reported that a key agenda of his meeting with TSMC executives would be securing a 'massive chunk' of 3nm production for AI products.

NVIDIA Could Account for a Major Portion of TSMC's 3nm Production, Intended for Rubin AI Lineup

It appears that the AI frenzy is in full force as of now, as with NVIDIA experiencing Blackwell Ultra ramp-up, the demand for chips has increased massively. According to a report by UDN, it is disclosed that one of Jensen's key reasons for visiting Taiwan is to tour TSMC's fab in Tainan, which is responsible for 3nm mass production. It is claimed that NVIDIA is negotiating an increase in capacity allocation for TSMC's N3 process to secure enough semiconductors to meet the anticipated demand around the next-gen Vera Rubin lineup.

Related Story Never-Released GeForce RTX 3050 Ti Desktop Card Appears Online, Featuring GA106 Die With 3328 CUDA Cores

The report claims that TSMC plans to expand the 3nm capacity of its fab located in Southern Taiwan Science Park from 100,000 wafers per month to 160,000 wafers, which is around a 50% increase, and interestingly, a large share of this capacity is expected to be allocated to NVIDIA 'exclusively'. It appears that NVIDIA is confident that the demand for Rubin AI chips is expected to be massive, which is why the firm has decided to reserve supply early on. For TSMC, this means that 3nm will drive the company's revenue for several quarters from hereon.

The Rubin AI lineup will be a 'pivotal' release from Team Green, considering that the architecture is expected to feature improvements all across the board, integrating a significant 'compute leap'. Apart from TSMC's N3P process, Rubin will feature the advanced HBM4 process, which will drive the lineup's capability to newer standards. Moreover, NVIDIA has already secured Rubin customers, despite almost two quarters remaining in mass production, which indicates that interest in the AI industry around the firm's new products is high.

For TSMC, HPC customers account for a significant portion of the firm's revenue, with NVIDIA being the primary source. The firm has begun integrating newer processes and plans to switch to A16 (1.6nm) technology relatively quickly. This is why, on each visit to Taiwan, Jensen Huang consistently emphasizes the importance of TSMC to his company.

Muhammad Zuhair Photo

About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.

Follow Wccftech on Google to get more of our news coverage in your feeds.

Button