Can Glide be run in parallel — for example on a cluster?
Although the underlying computational program (impact) is not parallelized, Glide docking jobs can be distributed over multiple processors. This is done by partitioning the ligands into sets and running each set on a separate processor. Beginning with Maestro 8.0, you can set up distributed docking jobs over multiple processors on multiple hosts in the Start dialog box. Simply specify the number of subjobs, select the hosts, and specify the number of processors on each host. You can also run distributed jobs from the command line with $SCHRODINGER/para_glide. The job can be set up in Maestro, and the input files written by clicking Write rather than Start. See the Glide User Manual for more information on para_glide. Note that each subjob requires a license. The number of processors chosen is therefore the maximum number of licenses that will be used. For instructions on configuring your Schrödinger installation for remote and batch queue submission, please see the “Preparing for Job S
Related Questions
- Can the Infrastructure (both Metadata Repository database and the Identity Management services) be run on a hardware cluster?
- Is there a version of PAUP* that will run a search in parallel on a multiple processor machine or a cluster of machines?
- How to construct and run a parallel job on the cluster?
