Unfortunately, Athena does not allow users to run programs while they are logged out ("batch jobs"), nor do we have a public grid computing cluster. If you have a program that requires a lot of time to run, there are a couple of options you have to consider:
First, see if your department can help you out. A professor with an Athena workstation on his or her desk may be willing to let you run a long job over the weekend, your department may have a private cluster where workstations can be reserved, or there may be other resources specific to your department.
You may also wish to try StarCluster, an open-source cluster-computing toolkit for use with the Amazon Elastic Compute Cloud (EC2). An EC2 account is required, but the software allows you easily stage a computing cluster and submit batch jobs. It includes popular software such as OpenMPI, ATLAS, Lapack, NumPy, SciPy, and you can customize the resulting master/slave images to your needs.
If neither of those options are suitable, you have some unofficial options:
- You can try using a workstation in Barker or Hayden, since the library closes at night. If you are able to obtain one of those workstations early enough to set up your job, be sure to turn on a screensaver when you leave. Note that normally, after twenty minutes, this allows any user to reboot you should he need the machine. If your job requires a longer run than overnight, you should get back to that workstation as soon as the library re-opens or you will be logged out.