New engine releases & news 2021

[Edsel Apostol]

I tried Invictus.r340.pext.exe. My hardware is Intel(R) Core(TM) i7-6950X CPU @ 3.00GHz 3000MHz, (10 cores, 20 threads)
I set the engine to play with 18 Threads and 512 MB Hash on Playchess server when this error occured with 2 cores less. Maybe there is a bug with MultiPV code?

Can you try opening the executable and run these commands:

setoption name Threads value 18
setoption name Hash value 512
go

You can also check the "invictus.log" for the last communication with the server

[Damir]

I have checked Invictus log but it seems is only an older version of Invictus.r323.pext there, not the latest r340.
I have uploaded the log from here: https://www.swisstransfer.com/d/db5f2df ... 9406a38574

I have tried and run the commands you told me. Except from the engine getting some knodes and Depth, there does not seem to happen that much...

[Werner]

I have had no problems on 1CPU.
And congrats to the huge jump:
1 Gogobello 2.2 x64 1CPU +45 +36/=41/-23 56.50% 56.5/100
2 Invictus r324 x64 1CPU -45 +23/=41/-36 43.50% 43.5/100

1 Invictus r340 x64 1CPU +147 +48/=44/-8 70.00% 70.0/100
2 Gogobello 2.2 x64 1CPU -147 +8/=44/-48 30.00% 30.0/100

[Edsel Apostol]

I have checked Invictus log but it seems is only an older version of Invictus.r323.pext there, not the latest r340.
I have uploaded the log from here: https://www.swisstransfer.com/d/db5f2df ... 9406a38574

I have tried and run the commands you told me. Except from the engine getting some knodes and Depth, there does not seem to happen that much...

The engine seems to work in your hardware. Maybe you have not configured it properly in your GUI. There should be log entries for r340 in there.

[Edsel Apostol]

I have had no problems on 1CPU.
And congrats to the huge jump:
1 Gogobello 2.2 x64 1CPU +45 +36/=41/-23 56.50% 56.5/100
2 Invictus r324 x64 1CPU -45 +23/=41/-36 43.50% 43.5/100

1 Invictus r340 x64 1CPU +147 +48/=44/-8 70.00% 70.0/100
2 Gogobello 2.2 x64 1CPU -147 +8/=44/-48 30.00% 30.0/100

Thanks Werner. 2 weeks of bugfixing and trying out some stuff makes a difference. Hope to see it beat Hannibal.

[Rebel]

I have checked Invictus log but it seems is only an older version of Invictus.r323.pext there, not the latest r340.
I have uploaded the log from here: https://www.swisstransfer.com/d/db5f2df ... 9406a38574

I have tried and run the commands you told me. Except from the engine getting some knodes and Depth, there does not seem to happen that much...

The engine seems to work in your hardware. Maybe you have not configured it properly in your GUI. There should be log entries for r340 in there.

Working flawlessly here.

Code: Select all

No. Engine              1     2     3     4     5     6  Score  Games   Perc   Moves
------------------------------------------------------------------------------------
 1 Invictus r340     xxxx  46.5  52.0  70.0  73.5  81.5  323.5 /  500 (64.70%)  64.4  
 2 Amoeba 3.3        53.5  xxxx   0.0   0.0   0.0   0.0   53.5 /  100 (53.50%)  63.3  
 3 Cheng 4.40        48.0   0.0  xxxx   0.0   0.0   0.0   48.0 /  100 (48.00%)  67.0  
 4 Drofa 3.0.0       30.0   0.0   0.0  xxxx   0.0   0.0   30.0 /  100 (30.00%)  57.3  
 5 Nemo 1.01         26.5   0.0   0.0   0.0  xxxx   0.0   26.5 /  100 (26.50%)  66.9  
 6 Cheese 2.2        18.5   0.0   0.0   0.0   0.0  xxxx   18.5 /  100 (18.50%)  67.4  

[amanjpro]

I tagged version 3.0 of Zahak, this comes with ~200 elo boost, thanks to the Texel optimization:

https://github.com/amanjpro/zahak/releases/tag/3.0

New features:

• Evaluation is tweaked with Texel optimization

• Better time management

• Rethinking mobility, aggressivity and king attack

• New binary for Apple Mac OSX M1

Which binary to use?

• Ubunut/Linux 64-bit (not Raspberry Pi): zahak-linux-amd64

• Raspberry Pi 3 family/Android: zahak-linux-arm32

• Raspberry Pi 4: zahak-linux-arm64

• Apple Mac OSX (64-bit, Intel): zahak-darwin-amd64

• Apple Mac OSX (M1 CPU): zahak-darwin-m1-arm64

• MS Windows 32-bit: zahak-windows-386

• MS Windows 64-bit: zahak-windows-amd64

The version naming convention is also changed to drop the fix part of the version name, i.e. instead of calling the version 3.0.0 we call it 3.0. To make version names easier for the end user.

Some match results:

Against Laser 0.1 (zahak_dev is version 3.0 and zahak_next is an intermediate step towards version 3.0):
Unfortunately, Laser kept disconnecting (namely in 17 games, on which 5 of them were drawn games, 4 of them were Laser was winning):

Code: Select all

Rank Name                          Elo     +/-   Games    Wins  Losses   Draws   Points   WWins  WLoss.  WDraws   BWins  BLoss.  BDraws
   0 zahak_dev                      98      29     400     198      88     114    255.0     111      33      56      87      55      58
   1 laser                         -10      40     200      66      72      62     97.0      41      29      30      25      43      32
   2 gopher_check                 -200      46     200      22     126      52     48.0      14      58      28       8      68      24

Finished match

Against Loki 3.5:

Code: Select all

Rank Name                          Elo     +/-   Games    Wins  Losses   Draws   Points   WWins  WLoss.  WDraws   BWins  BLoss.  BDraws
   0 Loki3.5                       129      29     400     210      68     122    271.0     117      28      55      93      40      67
   1 zahak_dev                    -102      41     200      40      97      63     71.5      22      42      36      18      55      27
   2 zahak_next                   -158      43     200      28     113      59     57.5      18      51      31      10      62      28

Finished match

Against Baislicka, Achillees, Clever Girl and Zahak 2.0.0

Code: Select all

Rank Name                          Elo     +/-   Games    Wins  Losses   Draws   Points   WWins  WLoss.  WDraws   BWins  BLoss.  BDraws
   0 zahak_dev                     194      24     800     519     114     167    602.5     278      50      72     241      64      95
   1 baislicka                    -143      44     200      36     114      50     61.0      20      53      27      16      61      23
   2 Achillees                    -193      49     200      32     133      35     49.5      15      67      18      17      66      17
   3 zahak-darwin-amd64-latest    -210      49     200      26     134      40     46.0      16      59      25      10      75      15
   4 clever-girl                  -235      50     200      20     138      42     41.0      13      62      25       7      76      17

Finished match

I suspect Zahak 3.0 to be around 2350-2380

[Damir]

Fire_8.N
Thanks to the Author FireFather Norman Schmidt

NNUE support
uci option enginemode: choose classic, nnue(default), or random

https://github.com/FireFather/fire/rele ... g/Fire_8.N

Releasd with couple of other binaries as well....

[kranium]

Fire_8.N
Thanks to the Author FireFather Norman Schmidt

NNUE support
uci option enginemode: choose classic, nnue(default), or random

https://github.com/FireFather/fire/rele ... g/Fire_8.N

Releasd with couple of other binaries as well....

Thx Damir

A couple notes:
To run this effectively, a processor that includes AVX2 (like BMI2) is almost mandatory, otherwise you could see an NPS decrease of up to 50%.
If you're not sure what instruction sets your CPU includes, I've included a complete of bmi2 and avx2 processors on the main page...see
https://github.com/FireFather/fire

ATM the nnue network must be named nn.bin and reside in the same directory as the Fire 8.N executable.

The NNUN implementation is completely different than SF, and utilizes a modified version of Daniel Shaw's/Cfish excellent nnue probe code:
https://github.com/dshawul/nnue-probe
Many thanks to him and Ronald de Man for that!

The download includes the top reinforcement learning NN (according to my extensive testing) trained by Sergio Viera last year.
I think you'll see that this, combined with Fire's search translates into some real interesting, unique, and strong engine chess.

[Damir]

DoctorB

V1.0.0

Thanks to the Author Hiblom slt

https://github.com/hiblom/DoctorB

DoctorB is a UCI compatible chess engine, written in C++.

DoctorB is named after a character in Schachnovelle by Stefan Zweig. Dr B plays chess against himself by seperating his mind in two personas: Black and White.

Current features
Classical bitboard board representation and move generation
Alpha beta search strategy with iterative deepening
Piece square board evaluation
Hashing / transposition table
Move ordering
Static Exchange Evaluation (SEE)
Repetition detection
Next up
Quiesence search
Support opening books
Threaded search to implement quit and stop commands properly
Down the road
Advanced board evaluation
End game tables
Aspiration Windows
Multi-threaded search