Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: hash numbers requested: authors please read

Author: Anthony Cozzie

Date: 15:29:05 09/18/02

Go up one level in this thread


Hardware: 1GHz Athlon Thunderbird. Hash Size: 200M. I bought RAM when it was
cheap about a year ago :-D
Hits: Keys matched, in table, depth was large enough to use.
Pings: Keys matched, in table.
Currently Zappa uses a single always-replace table.

Position #1:
1. Rf5 Qxa6 2. Qxa6 Rxa6 3. d7 Ra8 4. Rxc5 Rd8 5. Rc7
 = (1.45)       Depth: 8        00:00:07.52     2064kN

Time: 10008ms, Nodes: 2842k, Quiescence Nodes: 987k [35%]
Table Hits: 381k [13%], Table Pings: 516k [18%], NPS: 284k, Null Cuts: 38k

Note: Ironically, Zappa picks Re7 for every ply except 8! I don't know whether
its a "depth valley" or just a bug in the engine.

Position #2:
1. Rb7 Rb5 2. Rxb5 Kg8 3. Rb7 a5 4. Ra7 a4 5. Kg5 Kf8 6. Kxg4 a3
7. Kf3 a2 8. Rxa2 Ke7
 = (5.97)       Depth: 14       00:00:08.01     2879kN

Time: 10002ms, Nodes: 3956k, Quiescence Nodes: 468k [12%]
Table Hits: 2145k [54%], Table Pings: 2790k [71%], NPS: 396k, Null Cuts: 0k

Position #3
1. f4 Bg4 2. h3 Be6 3. f5 Bd7 4. fxg6 fxg6
 = (-0.09)      Depth: 8        00:00:06.40     1468kN

Time: 10081ms, Nodes: 2637k, Quiescence Nodes: 1175k [45%]
Table Hits: 321k [12%], Table Pings: 491k [19%], NPS: 262k, Null Cuts: 25k

Note: Zappa doesn't find Rxe7 even at high depths, preferring the f4-f5 plan out
to
depth 12 [3:00].



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.