Hi, sorry for the long wait for my answer, but I've been very busy. I'll try to be more exhaustive about my problem. My question is very simple: why is sbcl slower on a four core cpu (Intel Core duo quad) than on an old machine like my Intel pentium (single core).
I mean, my program runs on a single thread, but shouldn't be the single thread of the quad core cpu faster than the single thread of the single core cpu, being the latter at a lower frequency of the quad core? The fact is that the difference in speed is not of few seconds, but of 10 seconds versus the 2 minutes of the quad core.... The program is very simple, it uses only a small amount of memory. It takes an image in PGM format (200x200 pixels) and finds the edges of the objects in the image using an algorithm based on fuzzy logic. The percentage of CPU used on the single core CPU is 100% (for 10 seconds) and 25% on the quad core (this is as I expected... a single thread on a quad core cpu takes 1/4 of the CPU...) Well, that's all, I hope I've been more clear this time. (sorry for my bad english) Below you find some news about my sbcl version. Bye franZ >> CL-USER(7): (LISP-IMPLEMENTATION-VERSION) >> >> "1.0.29.11.debian" I use Lenny, my version is "1.0.18.debian" >> CL-USER(8): (not (not (member :SB-THREAD *features*))) >> >> T same result, like yours -- Linux Registered User: #414858 P Funking Band http://www.perugiafunkingband.it http://www.myspace.com/perugiafunkingband _______________________________________________ Gardeners mailing list [email protected] http://www.lispniks.com/mailman/listinfo/gardeners
