TSMC and NVIDIA Revolutionize Chip Manufacturing with cuLitho and AI

TSMC, a leader in semiconductor manufacturing, has announced its move to production using NVIDIA’s cuLitho computational lithography platform. This strategic collaboration aims to accelerate the manufacturing of advanced semiconductor chips, according to the NVIDIA Blog.

The Role of Computational Lithography

Computational lithography is a crucial process in transferring circuitry onto silicon, involving complex computations that include electromagnetic physics, photochemistry, and distributed computing. Historically, this step has been a bottleneck due to its compute-intensive nature, requiring vast data centers and consuming billions of CPU hours annually. A typical chip mask set can demand over 30 million CPU hours, making it a costly and time-consuming process.

Advancements with NVIDIA’s cuLitho

NVIDIA’s cuLitho platform introduces accelerated computing to this process, with 350 NVIDIA H100 Tensor Core GPU-based systems capable of replacing 40,000 CPU systems. This advancement dramatically reduces production time, costs, and resource consumption, enabling TSMC to push the limits of current semiconductor manufacturing capabilities.

Dr. C.C. Wei, CEO of TSMC, highlighted the integration of GPU-accelerated computing as a significant leap in performance, improving throughput, and reducing cycle time and power requirements. This development was discussed at the GTC conference earlier this year.

Generative AI Enhancements

Beyond accelerated computing, NVIDIA has integrated generative AI into the cuLitho platform. This integration enhances the creation of masks by delivering a 2x speedup in the optical proximity correction process. Generative AI aids in producing a near-perfect inverse mask, accounting for light diffraction, and expediting the process with traditional methods.

The combination of accelerated computing and AI is transforming semiconductor lithography, a field that has seen few rapid changes over the past three decades. These technologies enable more accurate simulations and realizations of complex mathematical techniques, previously hindered by resource limitations.

Implications for the Semiconductor Industry

The significant speedup in computational lithography accelerates the development of each mask in the fabrication process, reducing the overall cycle time for new technology nodes. With cuLitho, techniques like inverse lithography, once impractical due to time constraints, are now feasible, paving the way for the next generation of powerful semiconductors.

This collaboration between TSMC and NVIDIA marks a pivotal moment in semiconductor manufacturing, showcasing the potential of combining cutting-edge computing with AI to advance technology.

Image source: Shutterstock