Cadence leverages NVIDIA’s latest Blackwell system to increase the speed of solvers by up to 80 times;
Based on the new NVIDIA Llama Nemotron inference model, we are collaborating to develop a full stack proxy AI solution for engineering design and scientific applications;
Leading the way in adopting NVIDIA Omniverse Blueprint for AI factory digital twins, aimed at achieving efficient data center design and operation.
Shanghai, China, March 24, 2025- Cadence Electronics (NASDAQ: CDNS) announced today the expansion of its multi-year partnership with NVIDIA, with a focus on driving the development of accelerated computing and proxy AI. This cooperation aims to address key global technological challenges, promote innovative development in various industries, and achieve substantial breakthroughs.
Cadence drives technological innovation across multiple industries by integrating NVIDIA accelerated computing technology. Based on NVIDIA’s latest Blackwell architecture, Cadence? Series engineering and scientific solutions enable large-scale acceleration, enabling designers to tackle larger and more complex problems that were previously difficult to solve. The results of this collaboration are as follows:
The simulation time of computational fluid dynamics has been shortened by up to 80 times, from a few days to tens of minutes;
Cadence Spectre X Simulator accelerates up to 10 times;
The 3D-IC design and analysis of thermal, stress, and warpage can be accelerated up to 7 times.
Cadence Fidelity CFD platform utilizes Blackwell to solve the most complex fluid mechanics problems
Cadence utilizes the NVIDIA Blackwell platform to help solve a major challenge in computational fluid dynamics (CFD): simulating the entire aircraft in the most challenging parts of the flight envelope (during takeoff and landing). With the help of the Cadence Fidelity CFD platform, Cadence successfully ran simulations of billions of grid cells on NVIDIA GB200 GPU in less than 24 hours. Prior to this, the simulation process relied on a Top 500 CPU cluster with hundreds of thousands of cores and took several days to complete. Cadence will continue to utilize the limits of Blackwell test simulation to help the aerospace industry reduce wind tunnel testing frequency, lower costs, and accelerate the time to market process.
此外，Cadence 还与 NVIDIA 携手开发面向电子系统设计以及科学应用的全栈代理式 AI 解决方案。此次合作将引入突破性代理技术，将 Cadence JedAI 平台与 NVIDIA 的 NeMo 生成式 AI 框架和新发布的 NVIDIA Llama Nemotron 推理模型相整合，提高设计生产力，例如：
智能对话式 AI 助手，可提高用户生产力和创新能力；
基于底层设计资源和验证代理的深度验证推理；
利用设计代理，实现数字和定制电路的设计生成与优化。
In addition, Cadence Molecular Sciences (OpenEye) is integrating NVIDIA BioNeMo NIM microservices with Cadence’s cloud native molecular design platform Orion? Integration. This collaboration aims to accelerate the innovation of drug development tools by combining excellent cloud AI and GPU technologies. Orion has breakthrough on-demand and reserved GPU access capabilities, which can help global scientists carry out large-scale complex computing and revolutionize therapeutic drug design. NIM microservices have extended Orion’s capabilities, as follows:
An AI model used for heavy head prediction of protein three-dimensional structure;
Small molecule generative AI;
A basic AI model for predicting antibody characteristics.
Cadence is utilizing NVIDIA’s advanced digital twin technology to accelerate the construction of AI infrastructure. Cadence is proud to be one of the first companies to adopt NVIDIA Omniverse Blueprint, a digital twin designed for AI factories. This collaboration has promoted consistent and accurate model creation, enabling the rapid development of digital twins in data centers. NVIDIA Omniverse Viewport、Cadence Allegro? X Design Platform and Cadence Reality? The integration of Digital Twin Platform enables designers to gain a fresh perspective and more precise insight into the entire electronic system design process. Downstream users can utilize data for analysis and use in areas such as device and BOM management, manufacturing interfaces, system level quality, and mechanical and industrial design. NVIDIA and Cadence are at the forefront of creating high-quality model ecosystems, opening the door for device manufacturers and data center companies to quickly and confidently create digital twin solutions.
Cadence is accelerating AI driven EDA and system design and analysis workloads on NVIDIA’s latest Grace Blackwell NVL72 platform. We can help deliver infrastructure AI and agent-based AI that meet current needs, and reshape the simulation foundation for physical AI and scientific AI. ”Dr. Anirudh Devgan, President and CEO of Cadence, stated. With these technological breakthroughs, we are able to complete large-scale complex system simulations that were previously difficult to achieve within hours, including some of the largest and most accurate whole machine simulations to date
Accelerated computing and agent-based AI are redefining innovation standards in various industries. ”NVIDIA founder and CEO Huang Renxun said. NVIDIA and Cadence work together to continuously break through technological boundaries, making significant progress in simulation, optimization, and design, helping to improve efficiency, shorten product time to market, and drive scientific exploration to new heights
