Access a supercomputer for your NEURON modeling

NEURON code can easily be run in parallel (on many processors at once), saving you a large amount of time. This applies to both large-scale network simulations and very small simulations that you need to run hundreds of times to explore a parameterspace. To run NEURON in parallel, you need access to a computer with multiple processors or cores. Here are some ideas of how to find one.

1. How many cores does your personal computer have? Most computers, even laptops, have multiple cores in them, so even the little simulations you do on your own machine can be “parallelized”. Check the number of cores your computer has by doing one of the following:

  • Look at the task manager for your computer; it often has separate graphs for each processor
  • Research your computer make/model online for the specifications
  • On Windows, open a command prompt and enter:
    WMIC CPU Get NumberOfCores,NumberOfLogicalProcessors /Format:List

    or see other answers from this helpful StackExchange question.

  • On Linux, you can look at cpuinfo:
    cat /proc/cpuinfo | grep processor | wc -l

2. Are you affiliated with a school, company, or other institution? Your institution may have their own supercomputer (aka cluster computer) that you can likely access for free. Check with your institution’s information technology department.

3. Register for a free account with the Neuroscience Gateway portal (NSG). The NSG supports parallel execution of several neural simulators, including NEURON. It provides a user-friendly interface for the Trestles supercomputer (at San Diego Supercomputing Center), so that you don’t have to worry about any of the technical details of using NEURON on a supercomputer. With the NSG you could be running parallel NEURON on hundreds of processors today!

4. Apply for a supercomputer time grant. XSEDE, a program supported by the NSF, allows access to several very large supercomputers across the United States. Any professor (as well as postdocs) can apply for a startup allocation that will be reviewed and funded within a few weeks, with minimal paperwork required. Graduate students who have received an NSF GRFP or Honorable Mention are also eligible to apply. If the startup allocation works well, then you can apply for a full research allocation.

There are other programs similar to XSEDE that may also be worth checking out. Got other ideas? Post them in the comments!

Leave a Reply

Your email address will not be published. Required fields are marked *