Signal and Power Integrity (PCB/IC Packaging) Blogs

If you have ever had your car audio impacted by driving by power lines, you have experienced electromagnetic interference (EMI). If you have ever had to close your laptop for takeoff or landing, you’ve seen how serious the airline industry is about electromagnetic compliance (EMC). Electronic products need to make sure they meet government EMC requirements. The Aerospace and Defense industry needs to be sure that their devices are tolerant to various levels of EMI. For these reasons, an important part of electronic design is to meet the specification for EMC.

Over the last few quarters, Cadence and NVIDIA have worked on implementing a 3D-FDTD electromagnetic (EM) solver, Cadence Clarity 3D Transient Solver, on CUDA GPUs and using Message-Passing Interface (MPI) for multi-GPU simulation. The FDTD Solver is a transient, time domain solver wherein results for a wide frequency band can be obtained with one simulation as compared to frequency domain solvers that require multiple simulations for several predefined frequencies. The primary computational advantage with the Clarity Transient solver is that simulations can be easily parallelized across a single node (server) or across multiple nodes via partitioning the domain into different subdomains and solved efficiently and accurately using a heterogeneous mixture of CPUs and GPUs.

At the 2022 EMC Live Fundamentals event, Cadence and Nvidia will team up to show how electronic products can be designed in such a way that EMC can be verified before a design prototype is manufactured. 

To learn more and register for this event, please sign up here and select the webcast titled “Predicting System-Level Electromagnetic Noise and Immunity by Simulating using 3D FDTD Solver Technology with GPU Computing” scheduled for Tuesday, November 9, 2021 at 2:30 PM EST.

The webcast shows real world examples of how electronic product design cycles can be shortened by using an EMC simulation methodology from Cadence and GPU computing from Nvidia.