At my job I have access to a decently powerful cluster(right now 1048 cores, 25 nodes) that uses TORQUE with pbs scrips to submit jobs. Most of the time the cluster is running at under 50% load. Has anyone done any thing with a system like this before. I would like to submit jobs(single work units) to single nodes when they are not being used. I do not have admin privileges in my login.

Some of the nodes are single use nodes. The other nodes are for multi-node use(example,MPI).

I wouldn't know where to begin using such a powerful beast but my first reaction is that if you don't have admin privileges then you likely will not be able to install the required BOINC software and without that I doubt you will get far. I could be wrong so be sure to return to get follow up opinions/advice.

You might be able to get better/faster advice by asking at one or more of the projects whose admins are familiar to some extent with using a cluster. One such project that springs to mind is Einstein@home and there are probably others. I have a hunch they would bend over backwards to be able to use that cluster.

In the end I believe you'll have to get the system admin to install BOINC or something and that's where you'll hit the brick wall. Most admins see BOINC as a security risk (it isn't so bad, really) and they don't want to risk security and/or create more work for themselves on "their" system unless they get paid for it. Approach the people who own the system and ask them to tell the admin to install BOINC and deal with it.

I can run and install stuff it just has to be local in my home directory.

configure --prefix=$HOME/.local is what I use to compile programs.

It is more of the way we submit jobs to be run on the cluster does not fit the normal operation of the boinc client.

I will ask e@home about being able to help them. They are one of my normal projects anyhow.