Well, I guess that on a single processor machine, splitting a task into threads which essentially synchronously do the same thing each, is likely to give no improvement, since they end up competing for the same resources at the same time - the hog is the same, be it 1 hog or n times 1/n hogs. The picture changes when threads use different resources which can be consumed in parallel, or if more processing pipelines can be used (e.g. multiple cpus). The utility iostat() may give a picture of the scarcest resource.
--shmem
_($_=" "x(1<<5)."?\n".q·/)Oo. G°\ /
/\_¯/(q /
---------------------------- \__(m.====·.(_("always off the crowd"))."·
");sub _{s./.($e="'Itrs `mnsgdq Gdbj O`qkdq")=~y/"-y/#-z/;$e.e && print}