CFish

GM_2600 wrote:

Maybe. To find the app, enter in play google store and look for PLAY MAGNUS. YOU WILL FIND THE APP THERE.

Good morning .
Friend

Can we use Bin book in cfish for best performance blitz games against engine ??

Navneet detha wrote:

Good morning .
Friend

Can we use Bin book in cfish for best performance blitz games against engine ??

Yes u can..

Zippo wrote:

Navneet detha wrote:

Good morning .
Friend

Can we use Bin book in cfish for best performance blitz games against engine ??

Yes u can..

Which is best bin book now days??

Go to this site and see the ratibg of each books..

https://sites.google.com/site/computerschess/

Zippo wrote:

Go to this site and see the ratibg of each books..

https://sites.google.com/site/computerschess/

Please download link top 2 book ??

Please link perfect by Cevdet : 3715.

By ChessMan

Cfish 171220 x32-x64

Windows version

Updated to "AVX512, AVX2 and SSSE3 speedups". ...
syzygy1 committed 18 hours ago

Download: https://pixeldrain.com/u/CodZvemk

===========
It looks like versions without sparse are now faster than sparse for SSSE3_x32 (x64?), AVX2 and BMI2 assemblies. Possibly not on all processors

We need a test Sugar x Cfish

Waldemar wrote:

We need a test Sugar x Cfish

I do some test on android and SUGAR AI 1 IS WEAKER THAN CFISH 131220. CFISH IS RATED AROUND 3700, AND SUGAR AI 1.0 IS RATED AROUND 3600.

Cfish 131220 : rating 3710 and SUGAR AI 1 : rating 3630.

My phone is poco m2 pro ,please give a top engine ,

1) CICHESS 081220 ( 3720 )
2) CFISH 131220 ( 3710 )

1) CICHESS 081220 ( 3720 )
2) CFISH 131220 ( 3710 )

[quote="GM_2600"]1) CICHESS 081220 ( 3720 )
2) CFISH 131220 ( 3710 ) [/quote
Thank u🥰]

Cfish 191220 update for Android only (Compilation by me )
(This time used Termux + Arch-Linux + NDK )
aarch64-
https://www.dropbox.com/s/g8pxby3dz4ls8lc/Cfish-191220-aarch64.7z?dl=0
armv7-neon
https://www.dropbox.com/s/4puoaopmfdysc8n/Cfish-191220-arm7-neon.7z?dl=0
armv7-
https://www.dropbox.com/s/q9r5op3fyj62r97/Cfish-131220-armv7.7z?dl=0

I saw first time my build was faster compering to Skynet's builds atleast in my Device.

And I also run a 100 match tournament in CFA App.

As I excepted My Compilation won , Not sure in other devices my builds are compiled on my own phone maybe they are more optimised for my Phone's CPU.

My phone is poco m2 pro
Which link is better for me
aarch64,armv7-neon or armv7

anandhu002 wrote:

My phone is poco m2 pro
Which link is better for me
aarch64,armv7-neon or armv7

No second thought use aarch64.

Thank u very much for your fast msg

leonhurricaneX wrote:

Cfish 191220 update for Android only (Compilation by me )
(This time used Termux +  Arch-Linux + NDK )
aarch64-
https://www.dropbox.com/s/g8pxby3dz4ls8lc/Cfish-191220-aarch64.7z?dl=0
armv7-neon
https://www.dropbox.com/s/4puoaopmfdysc8n/Cfish-191220-arm7-neon.7z?dl=0
armv7-
https://www.dropbox.com/s/q9r5op3fyj62r97/Cfish-131220-armv7.7z?dl=0

I saw first time my build was faster compering to Skynet's builds atleast in my Device.

And I also run a 100 match tournament in CFA App.

As I excepted My Compilation won , Not sure in other devices my builds are compiled on my own phone maybe they are more optimised for my Phone's CPU.

Is this the same build ?? At same commit moment ? I see dates are different and total nodes also....
So no point in comparing.

For a comparison to be valid it has to be the same commit build. Total nodes has to be the same. Only time and nodes/sec can vary. Bench has to be done 20 or more times for each and compare the average.

Is this the same build ?? At same commit moment ? I see dates are different and total nodes also....
So no point in comparing.

Nope it's based on today's commit 16 hours ago by syzygy1.
https://github.com/syzygy1/Cfish.
Why do you think there is no point comparison my builds are compiled on my Phone and previous build 131220S are compiled on Desktop. First time Phones's build was faster comparing to desktop build.
And there is no big change in new commit.

chesstopengines wrote:

For a comparison to be valid it has to be the same commit build. Total nodes has to be the same. Only time and nodes/sec can vary. Bench has to be done 20 or more times for each and compare the average.

Sometimes try to use your brain I have run bench more than 20 times only the best one select from both of them, I know someone like you will not believe. So, I posted a tournament results for people like you.
Did you ever run a benchmark in your phone, I don't think so
Because you can't see the fixed depth is 16 for both of them and no one sets fixed number of nodes in benchmark.

leonhurricaneX wrote:

chesstopengines wrote:

For a comparison to be valid it has to be the same commit build. Total nodes has to be the same. Only time and nodes/sec can vary. Bench has to be done 20 or more times for each and compare the average.

Sometimes try to use your brain I have run bench more than 20 times only the best one select from both of them, I know someone like you will not believe. So, I posted a tournament results for people like you.
Did you ever run a benchmark in your phone, I don't think so
Because you can't see the fixed depth is 16 for both of them and no one  sets fixed number of nodes in benchmark.

It's not the best you have to take but the average.
As for bench, if it's internal bench of engine with same parameters and same net and same commit build it should be same total nodes.
I have nothing against you and it's Nice you compile for Android I am just saying that a comparison to be valid it has to be done like I said