Compute Nodes

The compute (batch) nodes at PDSF are heterogenous, reflecting the periodic procurement of new nodes (and the eventual retirement of old nodes).  From the user's perspective they are essentially all equivalent except that some have more memory per job slot.  If your jobs have memory requirements beyond the default maximum of 1.1GB you should specify that in your job submission and SGE will run your job on an appropriate node.  See Submitting jobs for more information.