NVIDIA Announces Expansion of CUDA-X Library

NVIDIA is expanding its chip design, verification, physical simulation, and data processing workflows into a GPU-accelerated software stack. In an announcement on July 26, NVIDIA revealed that it has expanded its agent toolkit and added Physics Nemo and the CUDA-X library in conjunction with DAC 2026. The newly added components include the iterative sparse solver cuISS, the direct sparse solver cuDSS, and cuEST for quantum chemistry. CUDA-X serves as a library layer that accelerates AI, data processing, high-performance computing, physical simulations, and quantum chemistry tasks, focusing on enabling AI agents to call physical models and simulation tools for design and verification. NVIDIA is collaborating with Cadence, Siemens, and Synopsys to develop engineering workflows that leverage its technology. Siemens is set to announce a verification-oriented agent AI workflow for semiconductor design, while Synopsys has unveiled an autonomous engineering workflow. Cloudera has announced native GPU acceleration for Apache Spark 4.1, broadening the scope of GPU acceleration through cuDF. NVIDIA's cuEST is currently in beta, and the commercialization of quantum chemistry calculations is still premature. The industry is paying more attention to the applicability of these features in real workflows rather than just the addition of functionalities.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.