Computer Chess Club Archives




Subject: branch misprediction

Author: Eric Oldre

Date: 08:30:22 07/14/04

Go up one level in this thread

On July 14, 2004 at 07:43:47, Dan Andersson wrote:

> Cache and other memory subsystems have little impact compared to IPC, BPU and
>branch misprediction penalty in next to all chess programs. Probably a fraction
>of a percent.
> Of the BSD distributions out there FreeBSD and OpenBSD have Tier 1 status on
>the AMD64 port. NetBSD port is fully funtional but won't be fully supported
>until next release.
>MvH Dan Andersson

I have heard many of the engine authors in this forum discussing the impacts of
branch (mis)prediction. As I don't have any real formal programming training and
and just beginning to learn C, I haven't really been able to understand.

a) what are branch mispredictions?
b) how do you minimize them?
c) how do you detect/measure them? (some sort of profiler i'm guessing).

No need to explain here on the board, but a link to a informational webpage
would be great.


This page took 0.01 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.