Synopsys Partners with NVIDIA to Revolutionize Chip Design with AI and Grace Blackwell
Synopsys Unleashes New Chip Design Power with NVIDIA Grace Blackwell
In a groundbreaking announcement made at GTC 2025, Synopsys, Inc., a leader in electronic design automation (EDA), revealed its collaboration with NVIDIA to push the limits of chip design technology. By optimizing its solutions for the NVIDIA Grace Blackwell architecture, Synopsys is set to enhance circuit simulations for next-generation semiconductors, boasting performance enhancements projected at up to 30 times faster.
Sassine Ghazi, the president and CEO of Synopsys, emphasized the urgency of optimizing chip design workflows, calling them some of the most complex engineering challenges. He remarked, "At GTC, we are unveiling the latest performance results from optimizing Synopsys solutions with NVIDIA's Blackwell platform, enabling truly revolutionary advancements in chip design that are essential for the productivity of engineering teams around the world."
Highlights of the Collaboration
The joint endeavor between Synopsys and NVIDIA focuses on an expansive range of EDA solutions that span design, verification, and manufacturing processes, all optimized for NVIDIA's powerful GPU technologies.
1. Circuit Simulation: The Synopsys PrimeSim™ tool, utilized for SPICE simulation needs, is projected to achieve considerable speed-ups of 30 times through the Grace Blackwell platform. This leap in performance will significantly reduce the time needed to sign off on challenging circuit design, transforming the process from days to mere hours.
2. Computational Lithography: The renowned Synopsys Proteus™, known for optical proximity correction (OPC), is also set to benefit from these advancements. By integrating NVIDIA’s technologies, the tool aims to realize speed increases of up to 20 times, facilitating faster production cycles at leading technology nodes.
3. Thermal and Electrical Analysis: The ambitious project also involves deploying NVIDIA's CUDA-X libraries in developing Synopsys tools like Sentaurus™. The expected outcome is an acceleration of TCAD process simulations, potentially achieving results that are 10 times quicker than current capabilities.
4. Materials Engineering: The QuantumATK® tool by Synopsys, integral for materials research and semiconductor analysis, intends to leverage the accelerated architecture for achieving up to 100 times faster results. This vastly improved efficiency can reshape the landscape of materials development in semiconductor manufacturing.
AI-Driven Design Solutions
On another front, Synopsys is making strides in enhancing chip design productivity with generative AI technology. Through the integration of NVIDIA NIM microservices into its knowledge assistant, Synopsys.ai Copilot, customers have reportedly experienced productivity improvements of twice their previous capabilities. The powerful combination of these tools promises a better time-to-information ratio, allowing engineers to solve complex problems faster.
Future Prospects
Looking ahead, Synopsys is committed to continuing this innovative journey by optimizing over 15 solutions to be compatible with NVIDIA’s Grace CPU architecture. This initiative is expected to further advance the capabilities of EDA workflows, including essential processes like physical verification and functional verification.
Synopsys will prominently showcase these cutting-edge solutions and advancements at GTC 2025, held from March 18th to 21st, where industry leaders gather to discuss the future of technology and innovation.
The impact of our collaborative efforts aims to empower semiconductor and systems industries across various applications. With this partnership, both companies are helping engineers unlock new levels of innovation and efficiency in chip design, paving the way for the next generation of technological advancement.