Amit k. Saha wrote: > On Wed, Jan 7, 2009 at 3:10 PM, Sibtey Mehdi <sibt...@infotechsw.com> wrote: >> >> I use multiprocessing to compare more then one set of files. >> >> The parallel processing time should be lesser. >> >> I am not able to get advantage of multiprocessing here. > > Depends on the overheads of spawning multiple threads/processes v$ > your actual processing. The same way, things are compared in old > school when they used to compare macros v$ functions in C/C++.
Simple file comparison is limited more by disk I/O than CPU. So by trying to read more than one file at a time, you are causing the disk to seek back and forth, losing time. Try it again but this time place the files to be compared in a RAM disk, and run both the multiprocess and single-process versions of your program. I'll best you see a difference then. Or spread them across multiple physical (not logical) disk drives. -Jeff _______________________________________________ BangPypers mailing list BangPypers@python.org http://mail.python.org/mailman/listinfo/bangpypers
