Computer Chess Club Archives




Subject: Internal Iterative Deepening

Author: Roberto Waldteufel

Date: 07:45:19 08/12/98

Hi all,

I have heard of the technique of internal iterative deepening, and I wonder if
anyone can explain the details. I already use iterative deepening at the root
(doesn't everyone?), and I am assuming that the "internal" refers to the same
thing at interior nodes of the main search (ie depth > 1). Does anybody know if
this works well with PVS/Negascout type algorithms, where most of the nodes are
searched with zero width? And what happens when you get fail high or fail low at
depths lower than the orriginal depth passed with the node? Do you ignore this
information and keep deepening, or are there cases when you use this added
weaker (but cheaper) information to selectively prune the node at the full
depth? Also, if anyone has done experiments with/without internal iterative
deepening, what was the order of magnitude of improvement it produced?

Best wishes,

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.