Using GPU for simulations in Ansys Lumerical FDTD™ can significantly speed up your simulations. For licensing purposes, the number of streaming multiprocessors (SMs) in a GPU is counted, like how the number of total cores is counted for the CPU solver.
This article explains how to find the number of streaming multiprocessors in your GPU within the product. The methods here are only applicable for local computers running FDTD. For remote nodes, please refer to external documentation such as those from Nvidia and third-party websites.
Resource configuration window
You can find the number of SMs in your GPU resource in the Configuration test section of the Resource Configuration manager. When you run tests, the window displays the number of SMs for your local GPU.
Log File
The FDTD log file shows the number of SMs near the top. This file is available after simulations and is in the same directory as your project file. It has the name <sim_name>_p0.log, where <sim_name> is your simulation file name.
All processes are communicating and ready to proceed to simulation...
CPU threads per process: 1
GPU streaming multiprocessors(SMs): 26
Ansys Lumerical 2025 R2.4 FDTD Solver Version 8.34.4336 (Windows 64bit)
...
Solver object result
The FDTD Solver object returns the total number of GPU SMs as a part of its results. You can find this result in the simulation benchmark section of the Result View.
Script access
The script command, gpuspecs returns the SM count of all GPUs in the system.
gpu_results=gpuspecs;
?gpu_results{1}.deviceSMCount;
result:
26See Also
Getting started with running FDTD on GPU, Resource configuration for single node GPU simulations, Running GPU simulations