Author: Richard A. Fowell (fowell@netcom.com)
Date: 20:52:50 01/15/99
I put speed benchmark data for HIARCS 6 in my CCR review of HIARCS 6 -
I'd like to do so for HIARCS 7 (with 65536K hash).
I'm especially interested in the following machines:
- Pentium II/450
- Pentium II/400
- Pentium II/333
- Pentium II/300
- Pentium Pro/200
- Mac G3/400
- Mac G3/300
- iMac
but I'm interested in other results, too - K6, Cyrix, etc.
To run this test, you'll need at least 70Mb of RAM and HIARCS 7.
It takes about 2.5 hours to run. Please email processed EPD files
to me at fowell@netcom.com.
I'll compile the results and post them here, and probably in a review
article, a little later.
======= instructions =======
Here's the instructions for the HIARCS 7 64Mb hash speed benchmark:
1) Make a copy of the test file "h7bench.epd" - call it
"myh7.epd", for example.
(HIARCS will write its results into the test file).
2) Set hash to 65,536K
(for Macintosh users - change the last line in the "hash" file to:
65536
Note that you need about 70Mb of free RAM for this to work -
Mac users can check the size of "largest unused block" in Finder
by looking in the Apple menu under "about this Macintosh")
3) Launch HIARCS 7.0 with the standard (as-shipped) setting.
(The easiest way is to copy a fresh copy from the CD
to your hard drive)
4) Verify the hash table setting in the
"HIARCS 7: hash table" menu setting
5) Set "Sound" off (to avoid annoyance)
6) Set "Level" to "Infinite"
7) Set "Options: Max Time" to 270 seconds
8) Use "File: Analyse EPD" on "myh7.epd".
9) <wait - this takes about 2.5 hours to run -
most convenient to set it up before going to bed/work>
10) Email a copy of the processed file to me,
(fowell@netcom.com) together with the OS version.
I'll compute the results and email them to you.
Background:
This EPD file is the LCTII test with CMB.02, CMB.03 omitted.
==========================================================================
NOTE: The "bm" moves below are all wrong! They have to be, for this test!
==========================================================================
The "best moves" (bm) in the .EPD file are actually "worst moves"!
If they were the realbest moves, HIARCS would find them, and stop thinking.
This would prevent the goal of measuring HIARCS nodes/second for a fixed time.
"bm" tags must be present to get HIARCS to record the node count,
but if HIARCS PV matches the "bm" tag, it quits thinking.
======= h7bench.epd =============
r3kb1r/3n1pp1/p6p/2pPp2q/Pp2N3/3B2PP/1PQ2P2/R3K2R w KQkq - bm Ng5; id "P.01";
1k1r3r/pp2qpp1/3b1n1p/3pNQ2/2pP1P2/2N1P3/PP4PP/1K1RR3 b - - bm Ng4; id "P.02";
r6k/pp4p1/2p1b3/3pP3/7q/P2B3r/1PP2Q1P/2K1R1R1 w - - bm Kb1; id "P.03";
1nr5/2rbkppp/p3p3/Np6/2PRPP2/8/PKP1B1PP/3R4 b - - bm Rb7; id "P.04";
2r2rk1/1p1bq3/p3p2p/3pPpp1/1P1Q4/P7/2P2PPP/2R1RBK1 b - - bm Qc5; id "P.05";
3r1bk1/p4ppp/Qp2p3/8/1P1B4/Pq2P1P1/2r2P1P/R3R1K1 b - - bm Qb1; id "P.06";
r1b2r1k/pp2q1pp/2p2p2/2p1n2N/4P3/1PNP2QP/1PP2RP1/5RK1 w - - bm Qg6; id "P.07";
r2qrnk1/pp3ppb/3b1n1p/1Pp1p3/2P1P2N/P5P1/1B1NQPBP/R4RK1 w - - bm Qh5; id "P.08";
5nk1/Q4bpp/5p2/8/P1n1PN2/q4P2/6PP/1R4K1 w - - bm Qd7; id "P.09";
r3k2r/3bbp1p/p1nppp2/5P2/1p1NP3/5NP1/PPPK3P/3R1B1R b kq - bm b3; id "P.10";
bn6/1q4n1/1p1p1kp1/2pPp1pp/1PP1P1P1/3N1P1P/4B1K1/2Q2N2 w - - bm Qf4+; id "P.11";
3r2k1/pp2npp1/2rqp2p/8/3PQ3/1BR3P1/PP3P1P/3R2K1 b - - bm Qe5; id "P.12";
1r2r1k1/4ppbp/B5p1/3P4/pp1qPB2/2n2Q1P/P4PP1/4RRK1 b - - bm Qd3; id "P.13";
r2qkb1r/1b3ppp/p3pn2/1p6/1n1P4/1BN2N2/PP2QPPP/R1BR2K1 w kq - bm Qe4; id "P.14";
1r4k1/1q2bp2/3p2p1/2pP4/p1N4R/2P2QP1/1P3PK1/8 w - - bm Qf6; id "C.01";
2k2b1r/1pq3p1/2p1pp2/p1n1PnNp/2P2B2/2N4P/PP2QPP1/3R2K1 w - - bm Qe4; id "C.04";
2r2r2/3qbpkp/p3n1p1/2ppP3/6Q1/1P1B3R/PBP3PP/5R1K w - - bm Qb4; id "C.05";
2r1k2r/2pn1pp1/1p3n1p/p3PP2/4q2B/P1P5/2Q1N1PP/R4RK1 w k - bm Qd3; id "C.06";
2rr2k1/1b3ppp/pb2p3/1p2P3/1P2BPnq/P1N3P1/1B2Q2P/R4R1K b - - bm Rc5; id "C.07";
2b1r1k1/r4ppp/p7/2pNP3/4Q3/q6P/2P2PP1/3RR1K1 w - - bm Qf5; id "C.08";
6k1/5p2/3P2p1/7n/3QPP2/7q/r2N3P/6RK b - - bm Qe3; id "C.09";
rq2rbk1/6p1/p2p2Pp/1p1Rn3/4PB2/6Q1/PPP1B3/2K3R1 w - - bm Qd3; id "C.10";
rnbq2k1/p1r2p1p/1p1p1Pp1/1BpPn1N1/P7/2P5/6PP/R1B1QRK1 w - - bm Bd7; id "C.11";
r2qrb1k/1p1b2p1/p2ppn1p/8/3NP3/1BN5/PPP3QP/1K3RR1 w - - bm Qg5; id "C.12";
8/1p3pp1/7p/5P1P/2k3P1/8/2K2P2/8 w - - bm Kb1; id "F.01";
8/pp2r1k1/2p1p3/3pP2p/1P1P1P1P/P5KR/8/8 w - - bm b5; id "F.02";
8/3p4/p1bk3p/Pp6/1Kp1PpPp/2P2P1P/2P5/5B2 b - - bm Bd5; id "F.03";
5k2/7R/4P2p/5K2/p1r2P1p/8/8/8 b - - bm Rc7; id "F.04";
6k1/6p1/7p/P1N5/1r3p2/7P/1b3PP1/3bR1K1 w - - bm Re2; id "F.05";
8/3b4/5k2/2pPnp2/1pP4N/pP1B2P1/P3K3/8 b - - bm Nf3; id "F.06";
6k1/4pp1p/3p2p1/P1pPb3/R7/1r2P1PP/3B1P2/6K1 w - - bm Ra3; id "F.07";
2k5/p7/Pp1p1b2/1P1P1p2/2P2P1p/3K3P/5B2/8 w - - bm Bc5; id "F.08";
8/5Bp1/4P3/6pP/1b1k1P2/5K2/8/8 w - - bm Kg2; id "F.09";
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.