Firenzina

Wed Feb 27, 2013 4:51 am by · Wed Feb 27, 2013 4:51 am by

http://dgusev.cs.edinboro.edu/GameDev/Chess/firenzina.html

Firenzina is a free, open-source UCI chess engine by Yuri Censor (Dmitri Gusev), a clone of Fire 2.2 xTreme by Kranium (Norman Schmidt).

Wed Feb 27, 2013 7:12 am by · Wed Feb 27, 2013 7:12 am by

Firenzina 2.2.2 says MSVCR110.dll is missing.
I found the file on the internet, and installed it into my windows, but I still get the error. sad

Wed Feb 27, 2013 7:18 am by · Wed Feb 27, 2013 7:18 am by

easl wrote:http://dgusev.cs.edinboro.edu/GameDev/Chess/firenzina.html

Firenzina is a free, open-source UCI chess engine by Yuri Censor (Dmitri Gusev), a clone of Fire 2.2 xTreme by Kranium (Norman Schmidt).

Thanks, what about Firenzina 64bit (not-popcnt) ?

Wed Feb 27, 2013 11:03 am by · Wed Feb 27, 2013 11:03 am by

unfortunately does not work for AMD ....

Wed Feb 27, 2013 1:15 pm by · Wed Feb 27, 2013 1:15 pm by

What's new in Firenzina 2.2.2?

Wed Feb 27, 2013 8:08 pm by · Wed Feb 27, 2013 8:08 pm by

Firenzina 2.2.2 xTreme x64 vs. Fire 2.2 xTreme x64
https://www.chess2u.com/t6908-firenzina-vs-fire#41350

Fri Mar 01, 2013 8:44 pm by · Fri Mar 01, 2013 8:44 pm by

Thomas Miller wrote:Firenzina 2.2.2 says MSVCR110.dll is missing.
I found the file on the internet, and installed it into my windows, but I still get the error.

Try this new MT compile that does not require DLLs from Microsoft Visual C++ 2012 Redistributable Package (x64):
http://dgusev.cs.edinboro.edu/GameDev/Chess/Firenzina_2-2-2_xTreme_popcnt_sse4-2_intel13_MT_x64.zip
I will release less restrictive compiles (no-popcnt, x32, etc) soon. I just need to test them as I make them.

Fri Mar 01, 2013 8:51 pm by · Fri Mar 01, 2013 8:51 pm by

Vladimir wrote:What's new in Firenzina 2.2.2?

It's a clone. New _mm_popcnt_u64 intrinsic replaced the old hardware popcnt, which, rather counter-intuitively, was slower than the software emulator of popcnt. The new hardware popcnt is faster, as it should be. Also, now you can have up to twice as many threads as CPUs. Not that you would want to, but you can.

Sat Mar 02, 2013 3:53 am by · Sat Mar 02, 2013 3:53 am by

Vladimir wrote:Firenzina 2.2.2 xTreme x64 vs. Fire 2.2 xTreme x64
https://www.chess2u.com/t6908-firenzina-vs-fire#41350

Your test must have been flawed. sad

There are several possibilities:
(a) You used Jose Velasco's no-frills Firenzina compile capable of running on my Intel Core 2 x64 Quad at work. (It doesn't support SSE4.2.)
(b) You forgot to set the number of threads to be the same for both engines. For Firenzina, set Max_Threads to be equal to the number of CPUs (say, 8 for a hyperthreaded Core i7 Quad).
(c) All of the above.
(d) You configured RobboBases for one engine, but not for the other one. (This is highly unlikely.)
I posted the results of the initial (02/25/13) Firenzina compile long time control 40/40 match against Stockfish 2.2.2 (sse42_ja_intel) at 6 threads for each engine at http://dgusev.cs.edinboro.edu/GameDev/Chess/FirenzinaMatch1vStockfish222_nunn_40-40.zip

Sat Mar 02, 2013 4:15 am by · Sat Mar 02, 2013 4:15 am by

Bob wrote:

easl wrote:http://dgusev.cs.edinboro.edu/GameDev/Chess/firenzina.html

Firenzina is a free, open-source UCI chess engine by Yuri Censor (Dmitri Gusev), a clone of Fire 2.2 xTreme by Kranium (Norman Schmidt).

Thanks, what about Firenzina 64bit (not-popcnt) ?

Will do, as well as a compile for x32, which others have requested.

Sat Mar 02, 2013 11:17 am by · Sat Mar 02, 2013 11:17 am by

@Yuri Censor wrote:

Vladimir wrote:Firenzina 2.2.2 xTreme x64 vs. Fire 2.2 xTreme x64
https://www.chess2u.com/t6908-firenzina-vs-fire#41350

Your test must have been flawed.

Are you kidding me? the test is 100% accurate/correct!

@Yuri Censor wrote:There are several possibilities:
(a) You used Jose Velasco's no-frills Firenzina compile capable of running on my Intel Core 2 x64 Quad at work. (It doesn't support SSE4.2.)

Your compile does not work, so i used his compile.

@Yuri Censor wrote:(b) You forgot to set the number of threads to be the same for both engines. For Firenzina, set Max_Threads to be equal to the number of CPUs (say, 8 for a hyperthreaded Core i7 Quad).

No.

@Yuri Censor wrote:(c) All of the above.

Very funny.

@Yuri Censor wrote:(d) You configured RobboBases for one engine, but not for the other one. (This is highly unlikely.)

No.

Sat Mar 02, 2013 5:07 pm by · Sat Mar 02, 2013 5:07 pm by

Vladimir wrote:

@Yuri Censor wrote:

Vladimir wrote:Firenzina 2.2.2 xTreme x64 vs. Fire 2.2 xTreme x64
https://www.chess2u.com/t6908-firenzina-vs-fire#41350

Your test must have been flawed.

Are you kidding me? the test is 100% accurate/correct!

@Yuri Censor wrote:There are several possibilities:
(a) You used Jose Velasco's no-frills Firenzina compile capable of running on my Intel Core 2 x64 Quad at work. (It doesn't support SSE4.2.)

Your compile does not work, so i used his compile.

That explains it, thank you. What equipment are you using? If your computer has a modern Intel processor that supports SSE4.2 and hardware popcnt, then you should be able to run my /MT compile that I released yesterday to remedy the problem. It's available at http://dgusev.cs.edinboro.edu/GameDev/Chess/Firenzina_2-2-2_xTreme_popcnt_sse4-2_intel13_MT_x64.zip
The initial /MD compile required DLLs from Microsoft Visual C++ 2012 Redistributable Package (x64). I overlooked it, because, as a developer, I have the DLLs, and I didn't realize that they're uncommon. redface

I ran a 500-game 1/1 match to quickly make sure that the /MT compile is no weaker than the initial /MD. The match ended in a tie today, +75-75=350, the kNodes/sec numbers were similar. The /MT executable is noticeably larger, due to static linking.

Compiler capabilities vary quite a bit, see some performance test results at http://software.intel.com/en-us/intel-parallel-studio-xe . Jose Velasco's is an MSVC compile.

Mon Mar 04, 2013 2:45 pm by · Mon Mar 04, 2013 2:45 pm by

Thanks for the new engine Yuri. Keep up the good work. smile

Sat Mar 09, 2013 1:26 pm by · Sat Mar 09, 2013 1:26 pm by

@Yuri Censor wrote:

Bob wrote:Thanks, what about Firenzina 64bit (not-popcnt) ?

Will do, as well as a compile for x32, which others have requested.

please link to Firenzina 64bit?

Sun Mar 10, 2013 3:59 am by · Sun Mar 10, 2013 3:59 am by

Bob wrote:

@Yuri Censor wrote:

Bob wrote:Thanks, what about Firenzina 64bit (not-popcnt) ?

Will do, as well as a compile for x32, which others have requested.

please link to Firenzina 64bit?

Firenzina 2.2.2 xTreme 64-bit, no popcnt: http://dgusev.cs.edinboro.edu/GameDev/Chess/Firenzina_2-2-2_xTreme_x64.zip

Sun Mar 10, 2013 6:33 am by · Sun Mar 10, 2013 6:33 am by

@Yuri Censor wrote:

Bob wrote:

@Yuri Censor wrote:

Bob wrote:Thanks, what about Firenzina 64bit (not-popcnt) ?

Will do, as well as a compile for x32, which others have requested.

please link to Firenzina 64bit?

Firenzina 2.2.2 xTreme 64-bit, no popcnt: http://dgusev.cs.edinboro.edu/GameDev/Chess/Firenzina_2-2-2_xTreme_x64.zip

64bit not 32bit, i have this console output:

Code:: Firenzina 2.2.2 xTreme w32 by Yuri Censor, a clone of Fire 2.2 xTreme by Kranium, based on Ippolit compiled by Yuri Censor Mar 6 2013 18:41:23

Sun Mar 10, 2013 7:24 am by · Sun Mar 10, 2013 7:24 am by

Bob wrote:

@Yuri Censor wrote:

Bob wrote:

@Yuri Censor wrote:

Bob wrote:Thanks, what about Firenzina 64bit (not-popcnt) ?

Will do, as well as a compile for x32, which others have requested.

please link to Firenzina 64bit?

Firenzina 2.2.2 xTreme 64-bit, no popcnt: http://dgusev.cs.edinboro.edu/GameDev/Chess/Firenzina_2-2-2_xTreme_x64.zip

64bit not 32bit, i have this console output:

Code:

Firenzina 2.2.2 xTreme w32 by Yuri Censor, a clone of Fire 2.2 xTreme by Kranium, based on Ippolit compiled by Yuri Censor Mar 6 2013 18:41:23

My sincere apologies and big thanks for helping catch the mistake! I was working on generic 32-bit and 64-bit compiles simultaneously and accidentally mixed them up. redface

Neither has been linked to the Firenzina website yet. Will fix.

Mon Mar 11, 2013 6:56 pm by · Mon Mar 11, 2013 6:56 pm by

Bob wrote:

@Yuri Censor wrote:

Bob wrote:

@Yuri Censor wrote:

Bob wrote:Thanks, what about Firenzina 64bit (not-popcnt) ?

Will do, as well as a compile for x32, which others have requested.

please link to Firenzina 64bit?

Firenzina 2.2.2 xTreme 64-bit, no popcnt: http://dgusev.cs.edinboro.edu/GameDev/Chess/Firenzina_2-2-2_xTreme_x64.zip

64bit not 32bit, i have this console output:

Code:

Firenzina 2.2.2 xTreme w32 by Yuri Censor, a clone of Fire 2.2 xTreme by Kranium, based on Ippolit compiled by Yuri Censor Mar 6 2013 18:41:23

Please try again! The link is the same,
http://dgusev.cs.edinboro.edu/GameDev/Chess/Firenzina_2-2-2_xTreme_x64.zip , but the archive contents changed. One 64-bit compile is LP, the other one is NLP (No Large Pages). Both are no-popcnt. I would like to know if they run okay on 64-bit AMD machines. (I don't have one.)

Mon Mar 11, 2013 7:06 pm by · Mon Mar 11, 2013 7:06 pm by

Thanks Yuri. thumb up

Mon Mar 11, 2013 10:01 pm by · Mon Mar 11, 2013 10:01 pm by

Code:: Games Completed = 1000 of 1000 (Avg game length = 15.287 sec) Settings = RR/32MB/2000ms+100ms/M 700000cp for 7 moves, D 150 moves/EPD:E:\LittleBlitzer\openings.epd(4000) Time = 7184 sec elapsed, 0 sec remaining 1. Firenzina 2.2.2 xTreme x64 510.0/1000 240-220-540 (L: m=220 t=0 i=0 a=0) (D: r=401 i=71 f=47 s=0 a=21) (tpm=113.7 d=13.57 nps=1574324) 2. Fire 2.2 xTreme x64 490.0/1000 220-240-540 (L: m=240 t=0 i=0 a=0) (D: r=401 i=71 f=47 s=0 a=21) (tpm=113.9 d=13.52 nps=1562418)

Mon Mar 11, 2013 10:25 pm by · Mon Mar 11, 2013 10:25 pm by

Bob wrote:

Code:

Games Completed = 1000 of 1000 (Avg game length = 15.287 sec) Settings = RR/32MB/2000ms+100ms/M 700000cp for 7 moves, D 150 moves/EPD:E:\LittleBlitzer\openings.epd(4000) Time = 7184 sec elapsed, 0 sec remaining 1. Firenzina 2.2.2 xTreme x64 510.0/1000 240-220-540 (L: m=220 t=0 i=0 a=0) (D: r=401 i=71 f=47 s=0 a=21) (tpm=113.7 d=13.57 nps=1574324) 2. Fire 2.2 xTreme x64 490.0/1000 220-240-540 (L: m=240 t=0 i=0 a=0) (D: r=401 i=71 f=47 s=0 a=21) (tpm=113.9 d=13.52 nps=1562418)

Thanks for testing! +7 Elo against the factory compile is not bad for a generic no-popcnt compile, where the new intrinsics don't play a role.

Tue Mar 12, 2013 6:43 pm by · Tue Mar 12, 2013 6:43 pm by

@Yuri Censor wrote:

Bob wrote:

Code:

Games Completed = 1000 of 1000 (Avg game length = 15.287 sec) Settings = RR/32MB/2000ms+100ms/M 700000cp for 7 moves, D 150 moves/EPD:E:\LittleBlitzer\openings.epd(4000) Time = 7184 sec elapsed, 0 sec remaining 1. Firenzina 2.2.2 xTreme x64 510.0/1000 240-220-540 (L: m=220 t=0 i=0 a=0) (D: r=401 i=71 f=47 s=0 a=21) (tpm=113.7 d=13.57 nps=1574324) 2. Fire 2.2 xTreme x64 490.0/1000 220-240-540 (L: m=240 t=0 i=0 a=0) (D: r=401 i=71 f=47 s=0 a=21) (tpm=113.9 d=13.52 nps=1562418)

Thanks for testing! +7 Elo against the factory compile is not bad for a generic no-popcnt compile, where the new intrinsics don't play a role.

I released this archive now by linking it to the Firenzina page, http://dgusev.cs.edinboro.edu/GameDev/Chess/firenzina.html

Tue Mar 12, 2013 6:56 pm by · Tue Mar 12, 2013 6:56 pm by

Compile Firenzina 2.2.2 for Mac OS X?

Thu Mar 14, 2013 5:31 pm by · Thu Mar 14, 2013 5:31 pm by

can we have a 32 bit version of the engine???

Thu Mar 14, 2013 6:12 pm by · Thu Mar 14, 2013 6:12 pm by

Victor wrote:Compile Firenzina 2.2.2 for Mac OS X?

I don't have a Mac, but if you or others are interested in doing that, you're most welcome!

by