Author: Robert Hyatt
Date: 10:29:22 01/27/00
Go up one level in this thread
On January 27, 2000 at 00:21:45, Ernst A. Heinz wrote: >>also remember that this is synchronous logic. The fast eval can give a quick >>exit, but it still takes 10 cycles to exit as I understand it. As he was very >>specific to say 24mhz processors searched 2.4 M nodes per second exactly. And >>his 20mhz procssors searched 2M nodes per second exactly. That tends to say >>that the fast eval/slow eval/other stuff are done in parallel and used as >>needed... It would be harder to design a piece of hardware that had a variable >>number of cycles per node without microprogramming the thing... > >At least in his IEEE Micro article he never says "exactly 10 cycles" >but he explicitly states "The average number of cycles per position >searched is about 10; ..." (page 75). > >Moreover, he repeatedly talks about lazy evaluation and fast exits >concerning the different parts of the static evaluation (pages 74-75). >His flowchart (Figure C, page 75) also gives sequential dependencies >for the "fast" and "slow" parts of the static evaluation, while >specifying parallelism for move generation and decision making (the >latter including both parts of the evaluation). > >=Ernst= There is certainly nothing preventing such a design, if the hardware is viewed as a finite state machine. It does complicate things... "the average" quote is pretty revealing, obviously... and does suggest some asynchronous exits are possible...
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.