Vajolet2 2.4 Release

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

elcabesa
Posts: 855
Joined: Sun May 23, 2010 1:32 pm

Vajolet2 2.4 Release

Post by elcabesa »

with pleasure I'd like to announce the release of Vajolet2 2.4

this release consist mostly of debugging and refactoring, I hope it adds few ELO point to the previous release.

I'd like to thank:
  • Dann Corbit, Jon Dart and Peter Osterlund for their help in find and solve multithread bottleneck.

    Jurgen Precour (LocutusOfPenguin) for the help with cmake and Raspberry build.

    Thomas Petzke, for his post on his "Standard Algebraic Notation Headache"

    Tcec boys for their beautyfull tournament, expecially Anton Mihailov .

    CCRL boys for their incessant work.
https://github.com/elcabesa/vajolet/rel ... jolet2_2.4

Marco
Damir
Posts: 2801
Joined: Mon Feb 11, 2008 3:53 pm
Location: Denmark
Full name: Damir Desevac

Re: Vajolet2 2.4 Release

Post by Damir »

Thanks Marco. I hope you keep improving your engine. :) :)
elcabesa
Posts: 855
Joined: Sun May 23, 2010 1:32 pm

Re: Vajolet2 2.4 Release

Post by elcabesa »

I'll try, but it's more and more difficult to improve it
LocutusOfPenguin
Posts: 34
Joined: Thu Sep 28, 2017 6:52 pm
Location: Karlsruhe, Germany
Full name: Jürgen Précour

Re: Vajolet2 2.4 Release

Post by LocutusOfPenguin »

Thanks Marco.
Just in time for my 0.9k release :D
elcabesa
Posts: 855
Joined: Sun May 23, 2010 1:32 pm

Re: Vajolet2 2.4 Release

Post by elcabesa »

LocutusOfPenguin wrote:Thanks Marco.
Just in time for my 0.9k release :D
is your program the core of the "DGT Pi - Chess Computer for DGT e-Boards"?
LocutusOfPenguin
Posts: 34
Joined: Thu Sep 28, 2017 6:52 pm
Location: Karlsruhe, Germany
Full name: Jürgen Précour

Re: Vajolet2 2.4 Release

Post by LocutusOfPenguin »

Hi Marco,
yes it is.

And with this upcoming release yr engine is part of it ....But this version is not "sold" right now...im always ahead , ha - but its part of auto-update.

Jürgen
fastgm
Posts: 818
Joined: Mon Aug 19, 2013 6:57 pm

Re: Vajolet2 2.4 Release

Post by fastgm »

Thank you for the new version, Marco!

I've tested it in my 60+0.6 sec. rating list http://www.fastgm.de
At this time control Vajolet2 2.4 had 422 time losses, in total of 15250 games, Vajolet2 2.3 only 24 time losses.

GUI is LittleBlitzer 2.74, default settings, Ponder false, OwnBook false
Maybe you can take a look at it.

Best regards,
Andreas

Code: Select all

Games Completed = 15250 of 15250 (Avg game length = 190.557 sec)
Settings = Gauntlet/128MB/60000ms+600ms/M 9000cp for 50 moves, D 500 moves/EPD:D:\Schach\LittleBlitzer-2.74\8moves_v3.epd(32000)
Time = 194537 sec elapsed, 0 sec remaining
 1.  Vajolet2 2.4             	7013.0/15250	4996-6220-4034  	(L: m=5798 t=422 i=0 a=0)	(D: r=2594 i=656 f=725 s=59 a=0)	(tpm=1236.8 d=23.01 nps=1479235)

Code: Select all

Games Completed = 13000 of 13000 (Avg game length = 189.859 sec)
Settings = Gauntlet/128MB/60000ms+600ms/M 9000cp for 50 moves, D 500 moves/EPD:D:\Schach\LittleBlitzer-2.74\8moves_v3.epd(32000)
Time = 165344 sec elapsed, 0 sec remaining
 1.  Vajolet2 2.3             	5267.0/13000	3504-5970-3526  	(L: m=5946 t=24 i=0 a=0)	(D: r=2069 i=603 f=799 s=55 a=0)	(tpm=1184.2 d=22.62 nps=1505297)
.
elcabesa
Posts: 855
Joined: Sun May 23, 2010 1:32 pm

Re: Vajolet2 2.4 Release

Post by elcabesa »

this result is really strange for me since i usually lety Vajolet play tournamet games at 6''+0.05'' without any timeout.

are you using your pc for only the games, or you are doing sone other works while the tournament is running? can you give me some more informations or debugging logs?
fastgm
Posts: 818
Joined: Mon Aug 19, 2013 6:57 pm

Re: Vajolet2 2.4 Release

Post by fastgm »

For this rating list I'm using a HP Z620 Workstation, Dual Intel Xeon E5-2670, 16 cores, 15 games are running at the same time, 1 core is free for Windows OS activities.
Of course, the workstation is not touched during all my tests. All other engines do not have a single time loss. That is the reason why I'm reporting this curious behavior.
Unfortunately, the LittleBlitzer GUI does not generate a log/debug file by default.

Code: Select all

Games Completed = 15250 of 15250 (Avg game length = 190.557 sec)
Settings = Gauntlet/128MB/60000ms+600ms/M 9000cp for 50 moves, D 500 moves/EPD:D:\Schach\LittleBlitzer-2.74\8moves_v3.epd(32000)
Time = 194537 sec elapsed, 0 sec remaining
 1.  Vajolet2 2.4             	7013.0/15250	4996-6220-4034  	(L: m=5798 t=422 i=0 a=0)	(D: r=2594 i=656 f=725 s=59 a=0)	(tpm=1236.8 d=23.01 nps=1479235)
 2.  Senpai 2.0               	177.0/250	140-36-74  	(L: m=36 t=0 i=0 a=0)	(D: r=39 i=18 f=17 s=0 a=0)	(tpm=1189.9 d=19.79 nps=1335894)
 3.  Sting SF 8.9 64bit       	153.0/250	106-50-94  	(L: m=50 t=0 i=0 a=0)	(D: r=63 i=15 f=14 s=2 a=0)	(tpm=1249.3 d=23.53 nps=1763198)
 4.  Arasan 20.3              	160.5/250	124-53-73  	(L: m=53 t=0 i=0 a=0)	(D: r=41 i=13 f=19 s=0 a=0)	(tpm=1201.0 d=20.72 nps=1217012)
 5.  The Baron 3.41 64-bits   	51.5/250	 23-170-57  	(L: m=170 t=0 i=0 a=0)	(D: r=41 i=7 f=9 s=0 a=0)	(tpm=726.5 d=15.05 nps=2268950)
 6.  Roc 0.8 x64              	209.0/250	179-11-60  	(L: m=11 t=0 i=0 a=0)	(D: r=37 i=5 f=17 s=1 a=0)	(tpm=1220.0 d=16.45 nps=1785122)
 7.  asmFish 171004 x64       	242.5/250	235-0-15  	 (L: m=0 t=0 i=0 a=0)	(D: r=9 i=2 f=4 s=0 a=0)	(tpm=1282.4 d=29.42 nps=2317620)
 8.  Cheese 1.9.1 64 bits     	37.5/250	 24-199-27  	(L: m=199 t=0 i=0 a=0)	(D: r=20 i=3 f=2 s=2 a=0)	(tpm=1250.8 d=15.09 nps=3328309)
 9.  Texel 1.07               	175.0/250	137-37-76  	(L: m=37 t=0 i=0 a=0)	(D: r=46 i=19 f=11 s=0 a=0)	(tpm=1245.2 d=15.99 nps=1762322)
10.  Alfil 15.8.16            	126.0/250	87-85-78      (L: m=85 t=0 i=0 a=0)	(D: r=55 i=13 f=9 s=1 a=0)	(tpm=964.4 d=15.43 nps=515112)
11.  Amoeba 2.6               	93.5/250	 53-116-81  	(L: m=116 t=0 i=0 a=0)	(D: r=66 i=8 f=7 s=0 a=0)	(tpm=1107.6 d=22.76 nps=1801425)
12.  Fire 6.1 x64             	231.0/250	213-1-36  	 (L: m=1 t=0 i=0 a=0)	(D: r=20 i=13 f=3 s=0 a=0)	(tpm=1318.4 d=18.35 nps=1862074)
13.  Houdini 6 x64            	241.5/250	233-0-17   	(L: m=0 t=0 i=0 a=0)	(D: r=12 i=5 f=0 s=0 a=0)	(tpm=1267.4 d=18.58 nps=1859587)
14.  Andscacs 0.92            	204.0/250	174-16-60  	(L: m=16 t=0 i=0 a=0)	(D: r=40 i=9 f=10 s=1 a=0)	(tpm=1281.0 d=21.07 nps=1352661)
15.  RuyDos 1.0.29            	39.0/250	 15-187-48  	(L: m=187 t=0 i=0 a=0)	(D: r=25 i=12 f=11 s=0 a=0)	(tpm=1343.6 d=17.56 nps=3136560)
16.  Demolito 2017-08-26      	99.0/250	 55-107-88  	(L: m=107 t=0 i=0 a=0)	(D: r=39 i=11 f=38 s=0 a=0)	(tpm=1133.9 d=18.99 nps=0)
17.  Ethereal 8.27            	47.0/250	 30-186-34  	(L: m=186 t=0 i=0 a=0)	(D: r=15 i=7 f=12 s=0 a=0)	(tpm=999.5 d=19.03 nps=2139245)
18.  Nirvanachess 2.4         	185.5/250	144-23-83  	(L: m=22 t=0 i=0 a=0)	(D: r=65 i=5 f=12 s=1 a=0)	(tpm=1191.6 d=20.25 nps=1510379)
19.  Komodo 11.2 64-bit       	239.5/250	232-3-15  	 (L: m=3 t=0 i=0 a=0)	(D: r=7 i=7 f=1 s=0 a=0)	(tpm=1359.1 d=23.17 nps=1748585)
20.  Pawny 1.2 x64            	41.0/250	 14-182-54  	(L: m=182 t=0 i=0 a=0)	(D: r=38 i=5 f=11 s=0 a=0)	(tpm=1332.4 d=18.35 nps=2276525)
21.  Nemorino 3.00            	151.0/250	101-49-100    (L: m=49 t=0 i=0 a=0)	(D: r=76 i=8 f=16 s=0 a=0)	(tpm=1216.6 d=21.15 nps=2101845)
22.  Pedone 1.6               	139.5/250	98-69-83  	 (L: m=65 t=0 i=0 a=0)	(D: r=47 i=19 f=15 s=2 a=0)	(tpm=1223.1 d=25.89 nps=2495951)
23.  Booot 6.2 x64            	190.0/250	151-21-78  	(L: m=21 t=0 i=0 a=0)	(D: r=54 i=16 f=7 s=1 a=0)	(tpm=1308.0 d=24.02 nps=1017099)
24.  Laser 1.4                	139.5/250	97-68-85  	 (L: m=68 t=0 i=0 a=0)	(D: r=46 i=22 f=15 s=2 a=0)	(tpm=1232.0 d=18.88 nps=1350972)
25.  Wasp 2.01                	112.5/250	64-89-97   	(L: m=89 t=0 i=0 a=0)	(D: r=54 i=17 f=23 s=3 a=0)	(tpm=1191.2 d=15.93 nps=1266303)
26.  Daydreamer 2.0.0-pre2    	81.0/250	 43-131-76  	(L: m=131 t=0 i=0 a=0)	(D: r=53 i=10 f=13 s=0 a=0)	(tpm=1235.7 d=25.99 nps=1234916)
27.  Rodent III 0.172         	91.0/250	 55-123-72  	(L: m=123 t=0 i=0 a=0)	(D: r=55 i=4 f=10 s=3 a=0)	(tpm=1179.1 d=18.46 nps=1518498)
28.  Tornado 8.0              	51.5/250	 25-172-53  	(L: m=172 t=0 i=0 a=0)	(D: r=31 i=9 f=5 s=8 a=0)	(tpm=1196.7 d=21.89 nps=1339250)
29.  Chiron 4 64bit           	200.0/250	166-16-68  	(L: m=16 t=0 i=0 a=0)	(D: r=46 i=9 f=13 s=0 a=0)	(tpm=1186.2 d=21.66 nps=1878941)
30.  Fizbo 1.9                	200.5/250	172-21-57  	(L: m=21 t=0 i=0 a=0)	(D: r=38 i=10 f=9 s=0 a=0)	(tpm=1353.2 d=17.52 nps=1469371)
31.  SmarThink 1.97           	149.0/250	116-68-66  	(L: m=68 t=0 i=0 a=0)	(D: r=43 i=5 f=18 s=0 a=0)	(tpm=1224.2 d=19.03 nps=1156221)
32.  Bobcat v8.0              	118.0/250	76-90-84  	 (L: m=90 t=0 i=0 a=0)	(D: r=49 i=9 f=26 s=0 a=0)	(tpm=1259.8 d=22.73 nps=2570629)
33.  Crafty 25.2              	112.0/250	73-99-78  	 (L: m=98 t=0 i=1 a=0)	(D: r=48 i=14 f=16 s=0 a=0)	(tpm=1292.5 d=22.16 nps=4846563)
34.  Deep Shredder 13 x64     	215.5/250	193-12-45  	(L: m=12 t=0 i=0 a=0)	(D: r=26 i=8 f=11 s=0 a=0)	(tpm=1276.7 d=20.83 nps=1878295)
35.  Stockfish 8 64           	239.5/250	230-1-19  	 (L: m=1 t=0 i=0 a=0)	(D: r=13 i=1 f=4 s=1 a=0)	(tpm=1325.1 d=28.77 nps=1767229)
36.  Fritz 15                 	199.0/250	159-11-80  	(L: m=11 t=0 i=0 a=0)	(D: r=50 i=17 f=13 s=0 a=0)	(tpm=1317.2 d=17.90 nps=803906)
37.  Hannibal 1.7 x64         	167.5/250	123-38-89  	(L: m=38 t=0 i=0 a=0)	(D: r=58 i=14 f=15 s=2 a=0)	(tpm=1130.0 d=23.40 nps=1741120)
38.  Hakkapeliitta 3.0 x64    	122.5/250	90-95-65  	 (L: m=95 t=0 i=0 a=0)	(D: r=47 i=4 f=11 s=3 a=0)	(tpm=1356.9 d=26.70 nps=3676743)
39.  iCE 3.0 v658 x64         	185.5/250	155-34-61  	(L: m=34 t=0 i=0 a=0)	(D: r=41 i=8 f=12 s=0 a=0)	(tpm=1175.5 d=31.50 nps=1718334)
40.  Cheng 4.39               	115.0/250	78-98-74  	 (L: m=98 t=0 i=0 a=0)	(D: r=51 i=7 f=14 s=2 a=0)	(tpm=1261.4 d=21.06 nps=2047086)
41.  Protector 1.9.0          	163.5/250	115-38-97  	(L: m=38 t=0 i=0 a=0)	(D: r=54 i=29 f=14 s=0 a=0)	(tpm=1246.7 d=15.90 nps=1280955)
42.  Deuterium v14.3.34.130   	78.5/250	 48-141-61  	(L: m=141 t=0 i=0 a=0)	(D: r=35 i=18 f=6 s=2 a=0)	(tpm=1278.3 d=17.03 nps=1261380)
43.  Mars 3.35 x64            	193.5/250	157-20-73  	(L: m=20 t=0 i=0 a=0)	(D: r=46 i=18 f=8 s=1 a=0)	(tpm=1276.7 d=18.42 nps=2630759)
44.  Equinox 3.30 x64mp       	185.0/250	141-21-88  	(L: m=21 t=0 i=0 a=0)	(D: r=56 i=13 f=19 s=0 a=0)	(tpm=1196.7 d=20.74 nps=1834749)
45.  Naum 4.6                 	155.5/250	119-58-73  	(L: m=58 t=0 i=0 a=0)	(D: r=39 i=20 f=14 s=0 a=0)	(tpm=1201.0 d=15.26 nps=2091361)
46.  Gull 3 x64               	197.0/250	163-19-68  	(L: m=19 t=0 i=0 a=0)	(D: r=43 i=16 f=9 s=0 a=0)	(tpm=1239.6 d=16.79 nps=2268785)
47.  Loop 2007 64-Bit         	55.5/250	 32-171-47  	(L: m=171 t=0 i=0 a=0)	(D: r=25 i=9 f=11 s=2 a=0)	(tpm=1310.7 d=14.40 nps=3313702)
48.  Jonny 4.00               	46.0/250	 24-182-44  	(L: m=182 t=0 i=0 a=0)	(D: r=35 i=3 f=6 s=0 a=0)	(tpm=1215.9 d=15.08 nps=1854739)
49.  Octochess r5190          	51.0/250	 27-175-48  	(L: m=175 t=0 i=0 a=0)	(D: r=34 i=6 f=8 s=0 a=0)	(tpm=1265.0 d=16.48 nps=1808837)
50.  Gaviota v1.0             	79.0/250	 49-141-60  	(L: m=141 t=0 i=0 a=0)	(D: r=44 i=11 f=4 s=1 a=0)	(tpm=1213.5 d=15.53 nps=864859)
51.  Murka 3 x64              	91.0/250	 45-113-92  	(L: m=113 t=0 i=0 a=0)	(D: r=61 i=18 f=8 s=5 a=0)	(tpm=1241.3 d=16.87 nps=3902685)
52.  MinkoChess 1.3 x64       	77.5/250	 45-140-65  	(L: m=140 t=0 i=0 a=0)	(D: r=42 i=13 f=10 s=0 a=0)	(tpm=1175.7 d=17.82 nps=1752710)
53.  Toga II 3.0              	66.0/250	 34-152-64  	(L: m=152 t=0 i=0 a=0)	(D: r=44 i=7 f=13 s=0 a=0)	(tpm=1266.8 d=16.83 nps=1202896)
54.  Quazar 0.4 x64           	102.5/250	72-117-61  	(L: m=117 t=0 i=0 a=0)	(D: r=38 i=4 f=18 s=1 a=0)	(tpm=1211.6 d=21.03 nps=713067)
55.  Critter 1.6a 64-bit      	197.5/250	158-13-79  	(L: m=13 t=0 i=0 a=0)	(D: r=55 i=6 f=17 s=1 a=0)	(tpm=1209.1 d=18.82 nps=3164120)
56.  Deep Rybka 4.1 x64       	181.0/250	142-30-78  	(L: m=30 t=0 i=0 a=0)	(D: r=52 i=10 f=13 s=3 a=0)	(tpm=1317.0 d=15.24 nps=148883)
57.  Deep HIARCS 14 WCSC      	130.0/250	90-80-80  	 (L: m=80 t=0 i=0 a=0)	(D: r=43 i=20 f=17 s=0 a=0)	(tpm=1198.4 d=18.79 nps=584127)
58.  Deep Junior Yokohama     	46.5/250	 26-183-41  	(L: m=183 t=0 i=0 a=0)	(D: r=30 i=6 f=3 s=2 a=0)	(tpm=1194.2 d=14.14 nps=2354247)
59.  Deep Sjeng c't 2010      	100.5/250	67-116-67  	(L: m=116 t=0 i=0 a=0)	(D: r=33 i=17 f=17 s=0 a=0)	(tpm=1264.1 d=15.05 nps=964371)
60.  Spark 1.0                	125.0/250	78-78-94  	 (L: m=78 t=0 i=0 a=0)	(D: r=70 i=13 f=10 s=1 a=0)	(tpm=1204.7 d=15.41 nps=2834775)
61.  Spike 1.4                	84.5/250	 45-126-79  	(L: m=126 t=0 i=0 a=0)	(D: r=68 i=5 f=5 s=1 a=0)	(tpm=1273.6 d=14.94 nps=1350096)
62.  DiscoCheck 5.2           	97.5/250	 60-115-75  	(L: m=115 t=0 i=0 a=0)	(D: r=43 i=6 f=22 s=4 a=0)	(tpm=1209.3 d=22.30 nps=0)
.
elcabesa
Posts: 855
Joined: Sun May 23, 2010 1:32 pm

Re: Vajolet2 2.4 Release

Post by elcabesa »

may I ask you to do some testing with a little few threads, for example 13 parallel games? just to understand what is happening with Vajolet?