Author: Eugene Nalimov
Date: 10:30:02 08/20/03
Go up one level in this thread
On August 20, 2003 at 03:59:38, Johan de Koning wrote: >On August 19, 2003 at 22:11:14, Robert Hyatt wrote: > >>On August 19, 2003 at 20:06:58, Mathieu Pagé wrote: >> >>>Hi, >>> >>>The fact: >>> >>>I have this question i read at some place that it is faster to unmake a move >>>than to save the state of the game before moving then restoring it when we want >>>to unmake the move. >>> >>>For the moment my engines did not implement unmake() (it is still buggy). >>> >>>My thougth: >>> >>>Since bitboard computation are slow (on 32 hardware) i think that it can be >>>slower to unmake the move than to save the state. I friend of me that is lot >>>better than me at optimizing code also think that. >>> >>>My questions: >>> >>>Are you all using unmake() function or there is some of you that found that >>>saving the state is better ? >> >> >> >>read the comments from Crafty in main.c. I started out using what is >>commonly called "copy/make" as that worked well in Cray Blitz. But it >>didn't work well in the PC. The PC has very limited memory bandwidth, >>when you compare the speed of memory to the speed/demands of current >>processors. If you keep the board in cache, and update it there, it is >>more efficient than to copy it from real memory to real memory... > >I hate to play Vincent here, but real memory is not an issue. > >If you manage to keep the deepest few plies worth of position structs in L1 >cache, then bandwith is pretty decent on the PC. And it has been ever since them >PCs were endowed with cache. > >Copying a struct does take time, and it can easily be pinpointed. Saving and >restoring and unupdating also takes time, but is harder to identify. Especially >since the stress on code cache and branch prediction don't show up in a run time >profile. > >... Johan Modern high-speed CPUs (PIII/P4/Athlon/AMD64/Itanium2, probably others, as I don't know their microarchitecture) are very memcpy-unfriendly. Search for "store forwarding" in your Pentium documentation, or use any search engine and search the Net. Very briefly: if you store some data (e.g. 32 bits), and then load only part of it (e.g. 8 bits), CPU will stall and wait till the data goes into cache, and then will reload it from there. As store buffers work in order that means that CPU will first flush earlier data into cache, and that can take quite some time because CPU can have up to 24 stores in flight... 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.