Perform parallel processing on Gaussian

To perform parallel processing in Gaussian, you can utilize Gaussian’s built-in capabilities to use multiple processor cores. This can significantly reduce quantum chemistry calculation times. The following explains how to set up parallel processing in Gaussian:

System Check and Initial Setup Ensure that your system supports multi-core processing and has the necessary resources (such as RAM and CPU).

Editing the Gaussian Input File To use parallel processing, you need to edit your Gaussian input file to specify the desired number of cores and the allocated memory. Example: Suppose your input file is named input.gjf. You need to add the keywords nproc and mem to the input file. %NProcShared=4 %Mem=4GB Your Gaussian route section here Title Card Required 0 1 Your molecular structure here In this example:

- %NProcShared=4 specifies that Gaussian should use 4 processor cores.

- %Mem=4GB determines the amount of RAM allocated to the calculations

Running Gaussian with Parallel Processing Settings After editing the input file, you can run Gaussian using the command line. Example: Bashg16 < input.gjf > output.log Here, input.gjf is the input file, and output.log is the output file where the calculation results are stored.
Settings in the Gaussian Configuration FileDefault.RouteDefault.Route%NProcShared=4 %Mem=4GB

Performance Evaluation After setting up parallel processing, you may need to evaluate the performance to ensure that the settings are optimal and your system is effectively using parallel processing. Notes:

1. Always check that your system has enough processor cores and sufficient RAM for parallel simulations.
2. Sometimes, using too many cores can lead to reduced performance due to processing overhead. Therefore, you may need to optimize the number of cores.

By following these steps, you can benefit from the parallel processing capabilities in Gaussian and reduce your computation time.

Request a supercomputer!

Perform parallel processing on Gaussian

Contact