Author: Sune Fischer
Date: 15:51:37 08/20/04
Go up one level in this thread
On August 20, 2004 at 17:52:58, Eugene Nalimov wrote: >On August 20, 2004 at 04:33:07, Tom Kerrigan wrote: > >>Now that AMD is selling two processors that are identical other than L2 cache >>size (Sempron has 256k, Athlon 64 has 512k) we have proof of Crafty's working >>set size: >> >>Sempron: 1,080,020 NPS >>Athlon 64: 1,080,230 NPS >> >>http://www.anandtech.com/linux/showdoc.aspx?i=2170&p=3 >> >>This should prove once and for all that Crafty's working set is < 256k and >>therefore that size of L2 cache has no effect on its performance (as long as >>it's >= 256k) and that main memory speed likely plays a trivial role >>performance-wise. >> >>I bring this up because of all of the long debates that have occurred in the >>past about the value of L2 cache, the speed of memory, and the working set size >>of chess programs. > >If you remember, several years ago I run the similar experiment. I benchmarked >Crafty on 900MHz Itanium2 with 1.5Mb of L2 cache, and on 900MHz Itanium2 with >3Mb of L2 cache. > >I don't remember exact numbers (I posted them here), but the 3Mb cache was >faster by (I believe) 10%. > >Unfortunately that did not answer original question, as CPUs have not only >different cache size, but also different cache associativity. 1.5Mb was 6-way >associative. 3Mb was 12-way associative. How big a hashtable did you use? If you used a very small hashtable and a relative large cache wouldn't you expect a significant part of the HT to get cached? I think it would, because usually when running with a very small hash the nps actually increase. If instead of decreasing the TH size one increased the cache size the effect would be equivalent I think. Perhaps it's best to turn off the hashtable entirely when testing for the "working set", otherwise there's no telling how hashing disturbs the measurement. -S. >Thanks, >Eugene
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.