it depends, you left out enough details that it could be correct. It would work that way if you are running 4 threads simultainiously but 1 thread being worked on by 4 processors doesn't happen.Bobo said:If I understand correctly, you can have 4 processors working simultaneously, and have "stuff" going through all 4 at the same time, and have it be 4 times as fast as 1 proc, but you can't really measure the speed. Is that correct?
I don't really agree with this, how would it be more efficient?Not faster, more efficient.
Cromewell said:, how would it be more efficient?
Are you sure about that? What if you had 1 thread running so 1 CPU was under full load and the other 3 are idle, are you trying to tell me thats efficient?Bobo said:Well it sure wouldn't be less efficient![]()
No, I really am notCromewell said:Are you sure about that?
What if you had 1 thread running so 1 CPU was under full load and the other 3 are idle, are you trying to tell me thats efficient?![]()
LOL he's right though ... also its kinda difficult to compare 4CPUx4task with 1CPUx1task because for the SMP solution it depends on memory requirements (which is why Opty's are so kewl cuz of their HT-based shared memory)lol smartass
LOL ya for server environments usually they have stuff coded to take advantage of it ... of course Amma_warrio wasnt talking about server environments (well if he was he wouldnt have "stumbled" here: http://www.computerforum.com/showthread.php?t=3061)Well, I hope that they designed it (if it really exists) so that it doesn't do that....
Praetor said:LOL he's right though