Hi folks, I'm working on a project for which the solution is highly parallelizable. I've been writing it so far for GHC as a single-threaded app. I'd like to be able to split the job into multiple pieces, and spawn different system threads for each piece, so they will run on separate CPUs. Either each thread needs to write to the same IO handle, or else the results need to be collected when the threads are merged, so the results can all be printed. I'm working on Linux. What's the easiest way to accomplish this? Glasgow Parallel Haskell? Is there a Posix.Threads library? I understand that Haskell threads (Control.Concurrent) all run in one system thread, so would not take advantage of SMP. I suppose in the worst case I can write a shell program that divides up the problem, runs several worker processes to output intermediate results to files, then combines the results to the desired filehandle. Thanks, Lyle