On Fri, 01 Sep 2006 16:35:08 -0500, <[EMAIL PROTECTED]> wrote: > At 02:59 PM 9/1/2006, you wrote: >>mos wrote: >> >>>AMD Athlon 64X2 3800+ Dual Core S939 Manchester (2x512K cache) >>>AMD Athlon 64X2 4200+ Dual Core S939 Manchester (2x512k cache) >>>AMD Athlon 64X2 4400+ Dual Core S939 Toledo (2x1MB cache) >>>AMD Athlon 64X2 4600+ Dual Core S939 Manchester (2x512k cache) >> >>I would think, as a blind guess, that the Toledo processor (twice the >>cache) would be the hands-down winner in this list. > > I just wanted to know if anyone actually has used the chip and has seen a > difference using 1mb cpu cache compared to the 512k cache. TIA
It would be meaningless unless you ran the identical programs. The classic problem with the big cache is too many out-of-cache hits causing cache dumps then reloads. In a small cache, that same problem takes less time to resolve. The problem gets ugly when various tricks (i.e. more lost time) get pulled to minimize the 'outs in a big cache. The short answer is that if big caches _really_ made a difference, every processor would have one. Specific applications can benefit, but the general result is bigger is not better. As for hyperthreading, my experience differs. -- Reply-To email is ignored. -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
