Local GROMACS Simulations Without Slowing Down Your Computer

Running molecular simulations can be demanding on your computer, especially when using software like GROMACS. If you’ve ever tried to run a local job and found your system nearly unusable during the simulation, you’re not alone. This is a common situation for molecular modelers who want to use their own hardware resources but also need their machines to stay responsive. Fortunately, the GROMACS Wizard in SAMSON offers an elegant way to balance performance and usability when running simulations locally.

When running GROMACS locally through the GROMACS Wizard, the software is designed to leave enough computing power for the rest of your operating system and applications to function normally. This is achieved through the default configuration, which uses less than the maximum number of threads. Let’s take a look at how this works and how you can tailor settings to fit your machine and workflow even better.

Customizing the number of threads

By default, GROMACS Wizard sets the number of threads for running local jobs to be less than the number of available logical cores on your machine. This means your simulation can run smoothly in the background without making the rest of your system unresponsive.

If you want to tweak this, head to the Settings menu in the GROMACS Wizard (click the gear icon at the top of the wizard interface):

Settings

Here’s what you can configure:

  • If Number of threads is set to 0, GROMACS will automatically use all available logical cores. This may lead to better performance, but your computer might become sluggish during runs.
  • If you set a non-zero value, say 4, and have 8 logical cores, then only 4 will be used, and the threads will be pinned to those cores using the -pin on flag. This improves performance by keeping threads on specific cores, while also preserving responsiveness.

Additional parameters for power users

Besides the thread configuration, the Settings menu lets you specify additional performance parameters such as:

  • ntmpi – number of MPI threads
  • ntomp – number of OpenMP threads
  • maxh – maximum runtime in hours
  • pme – Particle Mesh Ewald parameters

These advanced options allow you to fine-tune how GROMACS runs on your hardware. If you’re familiar with GROMACS or need to squeeze out every ounce of performance, this level of customization can be very helpful. For more technical descriptions of these parameters, refer to the GROMACS manual on performance.

Important scope note

Performance tuning options in GROMACS Wizard only apply to local jobs. If you’re submitting jobs to the cloud, these settings are not transferred. So, if your use case requires reproducible performance results on your machine, these settings are for you.

Quick tip: If your simulation seems unstable or is failing without a clear error, double-check the number of threads and pinning. Some simulations can behave differently depending on core allocation and system load.

To learn more about customizing local GROMACS performance in GROMACS Wizard, visit the full documentation page here.

SAMSON and all SAMSON Extensions are free for non-commercial use. You can get SAMSON at https://www.samson-connect.net.

Comments are closed.