FuryBench

FuryBench Testing Framework

home to Integral PlentyChess Obsidian Caissa Quanticade Serendipity Spaghet Schoenemann Clarity Illumina Pawnocchio Vine Anura

Active : 14 Machines / 432 Threads / 316.31 MNPS
Priority 0
DarkNe	Quanticade	net59	diff	10.0+0.10	LLR +1.28 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO +2.13 ± 2.50 (-0.38, +4.63) 95% GAMES 19,436 (24.9% 4,834W 50.9% 9,887D 24.3% 4,715L) PENTA [60⁻² 2,256⁻¹ 4,963⁰ 2,383⁺¹ 56⁺²]	net59 50% more training
Witek	Caissa	singular-exact-margin	diff	60.0+0.60	LLR +0.40 (-2.94, +2.94) bounds [-0.50, 2.50] elo ELO +1.02 ± 2.74 (-1.71, +3.76) 95% GAMES 14,940 (23.1% 3,453W 54.1% 8,078D 22.8% 3,409L) PENTA [13⁻² 1,733⁻¹ 3,927⁰ 1,791⁺¹ 6⁺²]
Anemat	Quanticade	more-hugepages	diff	4.0+0.04	LLR -1.31 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO -0.77 ± 2.65 (-3.42, +1.87) 95% GAMES 17,524 (24.9% 4,359W 50.0% 8,767D 25.1% 4,398L) PENTA [75⁻² 2,023⁻¹ 4,620⁰ 1,954⁺¹ 90⁺²]	huge pages for worker data
Priority -1
DarkNe	Quanticade	tuning	diff	20.0+0.20	TUNE 136 parameters ITER 1,918 / 10,000 GAMES 7,672 / 40,000	includes depths
Mergen	Illumina	neg-extensions	diff	8.0+0.08	LLR +0.00 (-2.25, +2.89) bounds [0.00, 5.00] elo ELO +0.00 ± 0.00 (+0.00, +0.00) 95% GAMES 0 (0.0% 0W 0.0% 0D 0.0% 0L) PENTA [0⁻² 0⁻¹ 0⁰ 0⁺¹ 0⁺²]	Negative extensions
Witek	Caissa	singular-beta-history	diff	10.0+0.10	LLR -1.57 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO -0.08 ± 1.23 (-1.31, +1.15) 95% GAMES 81,768 (23.6% 19,257W 52.9% 43,235D 23.6% 19,276L) PENTA [253⁻² 9,833⁻¹ 20,737⁰ 9,802⁺¹ 259⁺²]
Witek	Caissa	rfp-qsearch-verify	diff	10.0+0.10	LLR -2.32 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO -0.12 ± 1.04 (-1.16, +0.92) 95% GAMES 113,718 (23.5% 26,729W 53.0% 60,221D 23.5% 26,768L) PENTA [350⁻² 13,666⁻¹ 28,862⁰ 13,635⁺¹ 346⁺²]

Finished
Anemat	Quanticade	faster-conthist	diff	4.0+0.04	LLR -2.28 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO +0.08 ± 1.33 (-1.25, +1.41) 95% GAMES 68,042 (24.7% 16,831W 50.5% 34,395D 24.7% 16,816L) PENTA [357⁻² 7,415⁻¹ 18,443⁰ 7,468⁺¹ 338⁺²]	blah
Anemat	Quanticade	fmadd-better *	diff	4.0+0.04	LLR -2.29 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO -0.99 ± 2.15 (-3.14, +1.16) 95% GAMES 27,020 (24.9% 6,722W 50.0% 13,499D 25.2% 6,799L) PENTA [142⁻² 3,096⁻¹ 7,121⁰ 2,999⁺¹ 152⁺²]	weird idea
Anemat	Quanticade	prefetch-better	diff	4.0+0.04	LLR -2.52 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO -4.20 ± 3.43 (-7.63, -0.78) 95% GAMES 10,500 (24.1% 2,530W 50.6% 5,313D 25.3% 2,657L) PENTA [56⁻² 1,249⁻¹ 2,753⁰ 1,150⁺¹ 42⁺²]	Prefetch earlier, take 1
Anemat	Quanticade	mailbox-lover	diff	4.0+0.04	LLR -2.29 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO -1.70 ± 2.50 (-4.21, +0.80) 95% GAMES 18,964 (24.4% 4,636W 50.6% 9,599D 24.9% 4,729L) PENTA [87⁻² 2,125⁻¹ 5,157⁰ 2,020⁺¹ 93⁺²]	try mailbox improvements, idk
Anemat	Quanticade	reduce-memset	diff	4.0+0.04	LLR +2.99 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO +8.46 ± 3.92 (+4.53, +12.38) 95% GAMES 8,220 (25.9% 2,125W 50.7% 4,170D 23.4% 1,925L) PENTA [40⁻² 855⁻¹ 2,129⁰ 1,037⁺¹ 49⁺²]	blah
Anemat	Quanticade	pext-attacks	diff	4.0+0.04	LLR +2.98 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO +9.97 ± 4.36 (+5.61, +14.33) 95% GAMES 7,040 (26.6% 1,874W 49.6% 3,494D 23.8% 1,672L) PENTA [31⁻² 769⁻¹ 1,743⁰ 921⁺¹ 56⁺²]	pext attacks
Mergen	Illumina	skip-tt-in-se	diff	8.0+0.08	LLR -2.27 (-2.25, +2.89) bounds [0.00, 5.00] elo ELO -7.42 ± 6.34 (-13.76, -1.09) 95% GAMES 3,370 (20.4% 687W 57.1% 1,924D 22.5% 759L) PENTA [24⁻² 440⁻¹ 821⁰ 384⁺¹ 16⁺²]
Witek	Caissa	singular-exact-margin *	diff	10.0+0.10	LLR +2.96 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO +1.48 ± 1.13 (+0.34, +2.61) 95% GAMES 95,014 (23.8% 22,588W 52.9% 50,242D 23.3% 22,184L) PENTA [304⁻² 11,096⁻¹ 24,283⁰ 11,540⁺¹ 284⁺²]
Fury	Integral	tt_swar_16bit_key	diff	4.0+0.04	LLR -2.26 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -3.96 ± 3.23 (-7.19, -0.73) 95% GAMES 12,548 (25.2% 3,163W 48.4% 6,079D 26.3% 3,306L) PENTA [80⁻² 1,541⁻¹ 3,152⁰ 1,444⁺¹ 57⁺²]
Fury	Integral	tt_swar	diff	4.0+0.04	LLR -2.26 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -1.33 ± 2.14 (-3.47, +0.81) 95% GAMES 29,318 (25.4% 7,441W 48.9% 14,324D 25.8% 7,553L) PENTA [186⁻² 3,548⁻¹ 7,304⁰ 3,434⁺¹ 187⁺²]
DarkNe	Quanticade	[Obsidian] main	diff	10.0+0.10	ELO -47.78 ± 6.77 (-54.54, -41.05) 95% GAMES 2,810 of 10,000 (19.2% 539W 48.0% 1,348D 32.8% 923L) PENTA [28⁻² 523⁻¹ 669⁰ 175⁺¹ 10⁺²]	kekw this I expect to be bad
Fury	Integral	nnue_threat_inputs	diff	40.0+0.40	LLR +2.89 (-2.25, +2.89) bounds [0.00, 4.00] elo ELO +24.43 ± 7.77 (+16.68, +32.21) 95% GAMES 1,866 (28.2% 527W 50.5% 943D 21.2% 396L) PENTA [1⁻² 162⁻¹ 477⁰ 291⁺¹ 2⁺²]	Switch default network
DarkNe	Quanticade	tuning	diff	20.0+0.20	LLR +2.48 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO +5.69 ± 3.26 (+2.43, +8.96) 95% GAMES 9,158 (23.6% 2,161W 54.4% 4,986D 22.0% 2,011L) PENTA [2⁻² 883⁻¹ 2,658⁰ 1,035⁺¹ 1⁺²]	Be good please
Vast	Clarity	main	diff	40.0+0.40	ELO +25.96 ± 7.16 (+18.82, +33.12) 95% GAMES 2,534 of 7,500 (26.3% 666W 54.9% 1,391D 18.8% 477L) PENTA [12⁻² 228⁻¹ 591⁰ 431⁺¹ 5⁺²]	ltc core (pain)
Vast	Clarity	main	diff	8.0+0.08	ELO +16.50 ± 7.76 (+8.76, +24.25) 95% GAMES 2,466 of 10,000 (26.6% 655W 51.6% 1,273D 21.8% 538L) PENTA [26⁻² 237⁻¹ 587⁰ 360⁺¹ 23⁺²]	progtest because I'm bored and have spare hardware time
DarkNe	Quanticade	tuning	diff	20.0+0.20	TUNE 127 parameters ITER 7,230 / 10,000 GAMES 57,842 / 80,000	Tune duh
DarkNe	Quanticade	vnni	diff	2.0+0.02	LLR -2.30 (-2.25, +2.89) bounds [0.00, 2.00] elo ELO -0.49 ± 1.35 (-1.84, +0.86) 95% GAMES 72,214 (25.2% 18,178W 49.5% 35,756D 25.3% 18,280L) PENTA [558⁻² 8,070⁻¹ 18,904⁰ 8,066⁺¹ 509⁺²]	plz
Fury	Integral	vaporeon	diff	4.0+0.04	LLR +2.90 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO +3.27 ± 2.10 (+1.17, +5.37) 95% GAMES 26,660 (25.6% 6,831W 49.7% 13,249D 24.7% 6,580L) PENTA [131⁻² 2,721⁻¹ 7,377⁰ 2,968⁺¹ 133⁺²]
Anemat	Quanticade	vnni	diff	4.0+0.04	LLR -2.31 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO +0.50 ± 0.79 (-0.29, +1.29) 95% GAMES 185,466 (24.6% 45,532W 51.0% 94,669D 24.4% 45,265L) PENTA [788⁻² 19,965⁻¹ 50,919⁰ 20,314⁺¹ 747⁺²]	plz
Fury	Integral	[Obsidian] main	diff	40.0+0.40	ELO -4.31 ± 4.75 (-9.06, +0.44) 95% GAMES 5,000 of 5,000 (23.0% 1,151W 52.7% 2,636D 24.3% 1,213L) PENTA [12⁻² 600⁻¹ 1,330⁰ 554⁺¹ 4⁺²]
Witek	Caissa	quadruple_se_400	diff	10.0+0.10	LLR -2.95 (-2.94, +2.94) bounds [-3.00, 0.00] elo ELO -4.14 ± 2.60 (-6.74, -1.54) 95% GAMES 18,126 (22.9% 4,145W 53.1% 9,620D 24.1% 4,361L) PENTA [67⁻² 2,271⁻¹ 4,572⁰ 2,117⁺¹ 36⁺²]
Fury	Integral	ti-768-chef	diff	40.0+0.40	LLR +2.91 (-2.25, +2.89) bounds [0.00, 4.00] elo ELO +17.35 ± 6.50 (+10.86, +23.85) 95% GAMES 2,666 (26.9% 717W 51.2% 1,365D 21.9% 584L) PENTA [3⁻² 249⁻¹ 696⁰ 382⁺¹ 3⁺²]
Fury	Integral	ti-768-chef	diff	10.0+0.10	LLR +2.89 (-2.25, +2.89) bounds [0.00, 4.00] elo ELO +5.70 ± 3.58 (+2.12, +9.27) 95% GAMES 9,700 (26.3% 2,555W 49.0% 4,749D 24.7% 2,396L) PENTA [31⁻² 1,089⁻¹ 2,458⁰ 1,234⁺¹ 38⁺²]
Fury	Integral	ti-768-chef	diff	N=20000	LLR +2.94 (-2.25, +2.89) bounds [0.00, 4.00] elo ELO +10.15 ± 5.53 (+4.62, +15.68) 95% GAMES 6,714 (31.7% 2,129W 39.5% 2,652D 28.8% 1,933L) PENTA [158⁻² 729⁻¹ 1,417⁰ 865⁺¹ 188⁺²]
Witek	Caissa	eval-85-204B *	diff	10.0+0.10	LLR -0.96 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO +0.40 ± 0.76 (-0.35, +1.16) 95% GAMES 233,894 (23.0% 53,801W 54.1% 126,564D 22.9% 53,529L) PENTA [1,408⁻² 27,949⁻¹ 58,046⁰ 28,051⁺¹ 1,493⁺²]	Fix pext/pdep and MulHi64 falling back to software on non-Windows builds Also, with pext working, restore -DUSE_BMI2 in the makefile native probe. No functional change. Elo \| 3.01 +- 1.84 (95%) SPRT \| 5.0+0.05s Threads=1 Hash=8MB LLR \| 2.95 (-2.94, 2.94) [0.00, 2.00] Games \| N: 37274 W: 9297 L: 8974 D: 19003 Penta \| [203, 4093, 9721, 4418, 202]

1 2 3 220 221 222