New J.A Compiles

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

Moderator: Ras

User avatar
Aleksey52
Posts: 7
Joined: Fri Feb 20, 2026 5:11 pm
Full name: Aleksey Ivanov

Re: New J.A Compiles

Post by Aleksey52 »

Thank you, dear Jim, all Strelka 2.1 files worked on Android 7.1.1 in the DroidFish program in Computer vs Computer mode. Even the dot-prod architecture played, although previously the dot-prod engine was only defined, but could not be played.
However, I would still like version 5.0 - it should be stronger.
You once compiled Strelka 5.0 for computers and for arm5.
And if you still have the source code of version 5.0 in the archive, then it would be very cool to have these Strelka 5.0 compilations for Android.
[Google translator used, sorry for my English.]
Collector of chess engines for android.
User avatar
Jim Ablett
Posts: 2423
Joined: Fri Jul 14, 2006 7:56 am
Location: London, England
Full name: Jim Ablett

Re: New J.A Compiles

Post by Jim Ablett »

Aleksey52 wrote: Wed Feb 25, 2026 8:38 pm Thank you, dear Jim, all Strelka 2.1 files worked on Android 7.1.1 in the DroidFish program in Computer vs Computer mode. Even the dot-prod architecture played, although previously the dot-prod engine was only defined, but could not be played.
However, I would still like version 5.0 - it should be stronger.
You once compiled Strelka 5.0 for computers and for arm5.
And if you still have the source code of version 5.0 in the archive, then it would be very cool to have these Strelka 5.0 compilations for Android.
[Google translator used, sorry for my English.]
I don't have the source code to Strelka 5 any more.

Jim.
Jusuf
Posts: 6
Joined: Sat Aug 02, 2025 11:07 am
Full name: Udo Schlauch

Re: New J.A Compiles

Post by Jusuf »

Hi Jim,

is it possible to improve the official releases of:
a) Zangdar 6.1.1 (https://github.com/Carbecq/Zangdar/releases/tag/v6.1.1)
b) Minke 5.0 (https://github.com/enfmarinho/Minke/releases/tag/v5.0.0)

Thanks

Jusuf
SlonoSlonar
Posts: 31
Joined: Sat Feb 01, 2025 10:23 am
Full name: Pavel Podlozniy

Re: New J.A Compiles

Post by SlonoSlonar »

User avatar
Jim Ablett
Posts: 2423
Joined: Fri Jul 14, 2006 7:56 am
Location: London, England
Full name: Jim Ablett

Re: New J.A Compiles

Post by Jim Ablett »

Image
https://github.com/PGG106/Alexandria
Alexandria 9.0.0 by Guiseppe Pagano

Alexandria 9.0.0 64 JA

Windows/Linux & Android profiled builds

Proton link:
https://drive.proton.me/urls/G69EJKWZDM#NKfLk4faqHYA

Smash link:
https://fromsmash.com/Alexandria90JA

Jim.
Frank Quisinsky
Posts: 7284
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: New J.A Compiles

Post by Frank Quisinsky »

Code: Select all

   1 Stockfish 18 JA              :  3545.59   1300    83.4   870   428     2  1084.0   32.9  17.45  3190.23  12.07   26.0
   2 Stockfish 17.1 JA            :  3542.26   1300    83.2   862   438     0  1081.0   33.7  17.52  3190.23  12.07   26.0
   3 Reckless 260109 dev          :  3516.69   1300    81.3   814   486     0  1057.0   37.4  16.44  3190.23  12.07   26.0
   4 Uralochka 3.42a JA           :  3473.64   2350    79.3  1427   873    50  1863.5   37.1  12.46  3152.09  13.79   47.0
   5 Velvet 8.1.1                 :  3431.70   2350    75.9  1294   979    77  1783.5   41.7  12.06  3152.99  13.80   47.0
   6 CSTal 2.0                    :  3426.93   2350    75.5  1285   978    87  1774.0   41.6  11.79  3153.09  13.80   47.0
   7 Revenge 4.0                  :  3424.43   2350    75.3  1269  1000    81  1769.0   42.6  11.35  3153.14  13.81   47.0
   8 Igel 3.6.0 JA                :  3406.00   2350    73.7  1193  1077    80  1731.5   45.8  10.89  3153.53  13.82   47.0
   9 Zangdar 6.1                  :  3399.64   1300    71.3   632   590    78   927.0   45.4  14.24  3190.23  12.07   26.0
  10 Zangdar 6.1.1                :  3398.03   1300    71.2   618   614    68   925.0   47.2  14.65  3190.23  12.07   26.0
  11 Stockfish 200731 HCE dev     :  3377.28   2350    71.1  1147  1048   155  1671.0   44.6  10.87  3154.14  13.82   47.0
  12 SlowChess Blitz 2.9          :  3366.17   2350    70.1  1124  1046   180  1647.0   44.5  11.65  3154.38  13.81   47.0
  13 Texel 1.12                   :  3352.27   2350    68.8  1092  1049   209  1616.5   44.6  11.08  3154.68  13.82   47.0
  14 Monty 260208 dev             :  3333.92   1300    64.8   537   610   153   842.0   46.9  13.97  3190.23  12.07   26.0
  15 Minic 3.41                   :  3327.98   1300    64.2   491   686   123   834.0   52.8  14.44  3190.23  12.07   26.0
  16 Wasp 7.14 dev                :  3321.70   1300    63.5   529   593   178   825.5   45.6  13.89  3190.23  12.07   26.0
  17 Patricia 5 JA                :  3316.28   2350    65.3  1031  1009   310  1535.5   42.9  10.09  3155.44  13.84   47.0
  18 Tucano 12.00                 :  3315.11   1250    63.4   479   628   143   793.0   50.2  13.93  3182.85  11.69   25.0
  19 Wasp 7.00                    :  3313.01   2350    65.0  1044   968   338  1528.0   41.2  10.72  3155.51  13.83   47.0
  20 Tucano 12.00 JA              :  3295.17   1100    65.5   505   432   163   721.0   39.3  16.34  3124.35  16.11   22.0
  21 Stash 37.10 HCE JA dev       :  3262.47   1300    57.2   439   609   252   743.5   46.8  13.63  3190.23  12.07   26.0
  22 Nemorino 6.11 JA             :  3250.00   2350    58.7   879  1003   468  1380.5   42.7  10.45  3156.85  13.83   47.0
  23 Tcheran 9.0                  :  3220.09   2350    55.7   793  1032   525  1309.0   43.9  10.36  3157.49  13.83   47.0
  24 Leorik 3.1.3 JA              :  3213.84   2350    55.1   763  1062   525  1294.0   45.2  10.37  3157.62  13.83   47.0
  25 Prune 3.1.0 JA               :  3173.98   1300    47.6   327   583   390   618.5   44.8  13.55  3190.23  12.07   26.0
  26 Xiphos 0.6.1 HCE JA          :  3163.34   2350    49.9   687   971   692  1172.5   41.3  10.44  3158.70  13.83   47.0
  27 Booot 6.5 HCE                :  3159.38   2350    49.5   705   916   729  1163.0   39.0  10.16  3158.78  13.84   47.0
  28 Laser 1.7 HCE JA             :  3124.47   1250    42.3   281   495   474   528.5   39.6  13.70  3190.48  11.70   25.0
  29 Fizbo 2.0                    :  3092.89   1100    47.1   323   391   386   518.5   35.5  16.63  3124.35  16.11   22.0
  30 DanaSah 9.1 JA               :  3090.44   2350    42.5   535   928   887   999.0   39.5  10.98  3160.25  13.82   47.0
  31 Laser 1.7 HCE                :  3088.29   1100    46.7   311   406   383   514.0   36.9  16.54  3124.35  16.11   22.0
  32 Petrel 3.1 JA                :  3074.17   2350    40.9   514   894   942   961.0   38.0  10.93  3160.59  13.82   47.0
  33 Isa 4.5                      :  3065.72   1300    36.2   212   518   570   471.0   39.8  13.95  3190.23  12.07   26.0
  34 Senpai 3.0.2 HCE             :  3051.79   1300    34.8   210   486   604   453.0   37.4  13.88  3190.23  12.07   26.0
  35 Simbelmyne 1.10.0 HCE        :  3045.13   1300    34.2   215   459   626   444.5   35.3  14.58  3190.23  12.07   26.0
  36 Seredina 1.2c                :  3035.09   1300    33.2   183   498   619   432.0   38.3  14.40  3190.23  12.07   26.0
  37 Orion 1.1                    :  2998.92   1300    29.8   154   467   679   387.5   35.9  14.66  3190.23  12.07   26.0
  38 Demolito 251216 HCE          :  2983.22   1300    28.4   150   438   712   369.0   33.7  15.10  3190.23  12.07   26.0
  39 Fizbo 2.0 JA                 :  2982.97   1250    27.9   147   403   700   348.5   32.2  14.99  3196.14  11.65   25.0
  40 Critter 1.6a HCE             :  2975.25   2350    31.5   407   666  1277   740.0   28.3  11.04  3162.70  13.82   47.0
  41 Vajolet2 2.8.0 HCE           :  2965.31   2350    30.6   347   744  1259   719.0   31.7  11.14  3162.91  13.82   47.0
  42 Deep iCE 4.0.853 HCE JA      :  2950.06   1100    34.2   199   354   547   376.0   32.2  16.46  3124.35  16.11   22.0
  43 Deep iCE 4.0.853 HCE         :  2944.35   1250    24.4   121   368   761   305.0   29.4  15.77  3197.68  11.62   25.0
  44 Vault 0.4.1 JA               :  2890.22   1300    20.7   128   283   889   269.5   21.8  16.67  3190.23  12.07   26.0
  45 SmarThink 1.98 HCE           :  2886.70   2350    23.9   298   529  1523   562.5   22.5  12.16  3164.58  13.80   47.0
  46 Amoeba 3.4 HCE               :  2880.85   1300    20.0    87   347   866   260.5   26.7  16.74  3190.23  12.07   26.0
  47 Hakkapeliitta TCEC v2 HCE    :  2866.86   2350    22.4   262   528  1560   526.0   22.5  12.53  3165.00  13.79   47.0
  48 Pluto 1.0.1                  :  2834.10   1300    16.8    61   315   924   218.5   24.2  18.43  3190.23  12.07   26.0
  49 Spark 1.0 HCE                :  2766.42   2350    15.4   133   458  1759   362.0   19.5  14.27  3167.14  13.75   47.0
  50 Cheese 3.3.2 HCE             :  2758.69   1300    12.4    45   232  1023   161.0   17.8  19.63  3190.23  12.07   26.0
  51 KnightX 4.7 HCE              :  2701.40   1300     9.7    36   179  1085   125.5   13.8  22.36  3190.23  12.07   26.0
  52 Dumb 2.3 HCE                 :  2668.31   1300     8.3    33   150  1117   108.0   11.5  24.17  3190.23  12.07   26.0
Hi Jim,

something must be wrong with your Fizbo 2.0 compile.
But don't ask me what!

I made the same test with Laser.
Here your Laser compile is clearly better.

The third test is Deep iCE ...around the same.

The forth test is Tucano 12.00.
Here it is also a different to see.
But in case of Tucano, can be possible that on Intel the results are better.
If I remember ... Tucano is more optimated for Intel processors.

Only a smal test during my ETOC-G idea.

Best
Frank

Nice to see is the regular different (without Elo inflation) from Stockfish.
Current Stockfish 18 to the last version without NN.
Not more as 200 Elo, often I read here over 300 Elo.
Frank Quisinsky
Posts: 7284
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: New J.A Compiles

Post by Frank Quisinsky »

Hi Jim,

ah, I forgot ...
Your Laser 1.8 Beta JA ...
in my List called 1.7 HCE JA

I replayed my test-preparation-tourney results soon and will delete the Intel results.
Later I test your Laser 1.8 Beta JA as new engine.

I switched to Laser 1.7 HCE (official version) to one of my Standard-Test-Engines.
For the moment all is not very interesting with this list.

But fact is ...
Something is wrong with your Fizbo 2.0 compiles.
Also your last Minic compiles lost around 85-90% of games on time.

Best
Frank

Bad is ...
Didn't achieve my own goal.
I actually wanted to test all TOP-275 engines with a draw quote not exceeding 50% with a balanced book. Oh well, I can't predict the future when it comes to selecting "Standard-Engines". Nothing is perfect, but I am in the near with the draw-quote to avoid Elo-Inflation.
PaulieD
Posts: 245
Joined: Tue Jun 25, 2013 8:19 pm

Re: New J.A Compiles

Post by PaulieD »

Reckless 0.9.0 is out, please compile
https://github.com/codedeliveryservice/ ... tag/v0.9.0
Jouni
Posts: 3839
Joined: Wed Mar 08, 2006 8:15 pm
Full name: Jouni Uski

Re: New J.A Compiles

Post by Jouni »

Note about Alexandria. Official compile has bench 9 600 514 nodes, but JA compile 9 276 730 nodes.
Jouni
MirceaH
Posts: 51
Joined: Mon Mar 13, 2006 4:21 pm

Re: New J.A Compiles

Post by MirceaH »

@jouni: Download nn.net and bench will report correctly