Author: Robert Hyatt
Date: 10:16:54 01/15/98
Go up one level in this thread
On January 15, 1998 at 10:49:05, Amir Ban wrote: >On January 15, 1998 at 10:27:28, Stuart Cracraft wrote: > >>Also, I think we need to discuss rehashing. Does somenoe >>have this implemented? It seems like rehashing for say >>8 slots based on nodes-recorded would be better than >>single-tier since quite a few collisions could be handled >>without replacing information already at that slot. >> >>Does anyone have some good algorithms for this? Let's say >>I hash to index X and find it is filled. What is a good >>algorithm to choose another index to try? And it seems like >>that algorithm would have to be deterministic because the >>next time into this entry, I'd want to go search those >>alternate indexes/slots in the same order to find where I'd >>stored it. >> >>--Stuart > >This is called "multiple-probes". Just continue sequentially from the >original hash location for as many probes you decide to do. > >Amir This is actually the best way to do it on a PeeCee. Because when you probe to a random address, you fetch 32 bytes into a single line of cache. On a non-PC, it is better to use a non-constant offset that is obtained from another part of the hash key, based on lots of experiments run a few years ago..
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.