Author: Dann Corbit
Date: 14:25:01 11/16/01
Go up one level in this thread
On November 16, 2001 at 17:15:45, Will Singleton wrote: >On November 16, 2001 at 15:00:53, Dan Andersson wrote: > >>It seems that the AltiVec engine is not working OK in the speedier Motorola >>chips. Until that is corrected, the question is moot. >> >>MvH Dan Andersson > >Well, let's assume they get it working. Chess doesn't seem to lend itself well >to a small vector unit, at least in traditional current algorithms. Do you know >enough about vector programming to opine on that? If we have a vector of computation registers, I think that would be a very good chess engine, especially if they can run truly parallel. Suppose (for instance) that I have 256 registers. Each register can work on one of the subtrees created by the move generator. If it's just a dot product thingy, I doubt if it would help any.
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.