Fire 2.2 xTreme

@stockfishJA wrote:

dear sir,

could you please elaborate on what Verbose_robbobases is?

thank you in adv.

stock

I was also wondering about that. But actually I'm more or less wondering about *all* the parameters available in Fire. It's nice to have the option to tweak your own version, but if you have no idea what the parameters stand for it's best not to touch them...

Is there any source where one can find out what the different options mean and how changes would affect the engine? And which of the parameters that are connected and need to be aligned with each other?

please, i have a question:
if i want that fire doesnt do draws, which is the number 1 or 200?
sorry my english
happy chess!!

Seems there should be more explanation for each parameter, this will be very useful for Fire users

It should be available what each parameter means and how it exchanges functtions with others....

i did prepare such a document for the release of Fire 1.31 -> UCI options.rtf
it's included in the 1.31 download archive.

it has a detailed description of all Fire 1.31 UCI options, many of which are still present in Fire 2.2

at the moment i am taking a break, to recharge batteries, so to speak...
i do intend, however to revise and expand the 1.31 UCI options.rtf to include all the the new Fire 2.2 options
in the meantime, this existing doc might be a good start.

Here's a link in case you don't have it:
http://www.chesslogik.com/downloads/Fire_131_x64_KLO_popcnt.rar

Hi kranium,
what different between :
Fire_22_xTreme_x64_PP and Fire_22_xTreme_x64_NLP_PP
thanks

regards :D

@kudatrojan wrote:

Hi kranium,
what different between :
Fire_22_xTreme_x64_PP and Fire_22_xTreme_x64_NLP_PP
thanks

regards :D

Just a guess: NLP = no large pages

@Richard Vida wrote:

@kudatrojan wrote:

Hi kranium,
what different between :
Fire_22_xTreme_x64_PP and Fire_22_xTreme_x64_NLP_PP
thanks

regards :D

Just a guess: NLP = no large pages

thanks Richard same guess to me.

and to make sure : is this right kranium ???

@kudatrojan wrote:

@Richard Vida wrote:

@kudatrojan wrote:

Hi kranium,
what different between :
Fire_22_xTreme_x64_PP and Fire_22_xTreme_x64_NLP_PP
thanks

regards :D

Just a guess: NLP = no large pages

thanks Richard same guess to me.

and to make sure : is this right kranium ???

most already know this...as an explanation is on the website:
http://www.chesslogik.com/fire.htm

and in the documentation.

large pages does increase the playing strength in my computer

In order for the large page to be working you have to do the fallowing:

Large pages (huge pages) can increase performance by up to 15 percent.
Windows
1. Choose Start > Control Panel > Administrative Tools > Local Security Policy.
2. In the left pane of the Local Security Settings window, expand Local Policies and choose User Rights Assignment.
3. In the right pane of the Local Security Settings window, choose Lock pages in memory and choose Action Properties. The Local Security Setting dialog box opens.
4. In the Local Security Setting dialog box, click Add User or Group.
5. Enter the appropriate user name, then click OK to close the Select Users or Groups dialog box.
6. Click OK to close the Local Security Policy Setting dialog box. The Local Security Settings window shows the user assigned to the policy.

its not new its the " No Large Pages " compile

Where are the settings ?

Haven't see any miracle settings yet but if you post them I'll give it a try. On my 3 minutes test Fire 2.GH Cremator1 did go over 36% on 50 games. https://www.chess2u.com/t3515-fire-22-gh-cremator1-tests#21021
I'll give you the benefit of the doubt but we need to see those settings first.

i think this fire can not win houdini 1.5a not matter wich setings but more than draw can not i tested in 3/5 15/10 10/25 blitx but hahah houdini is stronger

I guess there will always be someone trying to kill the King, or at least conspiring to do it. Its normal human nature.

@eddy158k wrote:

I guess there will always be someone trying to kill the King, or at least conspiring to do it. Its normal human nature.

@eddy158k wrote:

Haven't see any miracle settings yet but if you post them I'll give it a try. On my 3 minutes test Fire 2.GH Cremator1 did go over 36% on 50 games. https://www.chess2u.com/t3515-fire-22-gh-cremator1-tests#21021
I'll give you the benefit of the doubt but we need to see those settings first.

+1

@prapqi wrote:

i think this fire can not win houdini 1.5a not matter wich setings but more than draw can not i tested in 3/5 15/10 10/25 blitx but hahah houdini is stronger

+1

Houdini is much stronger than fire 2.2.
The Strength of an engine mainly depends on main cores of lines of codes, search methods and evaluations.
People trying to improve strength with the setting alone is like alchemists in centuries ago dreaming of changing from brass to gold.

houdini and fire i use us a engine

help pls.
i play
my softwarei setup never resign sometime stop i mean not moving, i click a lot move now stil not moving, i setup never resign

I know which one will beat Houdini 1.5a..... Houdini 2.0

I can't wait!

My new Fire 2.2 Blaze settings look very promising. Here is the current standings of my 2+2 tourney, in Arena, calculated with EloStat:

Program Elo + - Games Score Av.Op. Draws

1 Fire 2.2 Blaze 2882 69 68 47 58.5 % 2822 53.2 %
2 Houdini 1.5 2877 30 30 274 62.4 % 2789 47.4 %
3 Fire 2.2 Inf-DTU2 2823 44 44 102 51.5 % 2813 57.8 %
4 DeepSaros 2.3i 2822 48 48 68 50.7 % 2817 66.2 %
5 Stockfish 2.1.1 2820 48 48 111 51.4 % 2810 45.0 %
6 Rybka 4 2812 45 45 125 52.0 % 2798 46.4 %
7 Fire 2.2 Cremator 2811 42 42 99 48.5 % 2822 62.6 %
8 Critter 1.2 2805 38 38 135 47.4 % 2823 57.8 %
9 Leopard 07c-P 2805 43 44 82 48.2 % 2817 67.1 %
10 IvanHoe 47c GH 2792 39 40 102 46.6 % 2816 65.7 %
11 Fire 2.2 xTreme 2766 53 55 76 38.2 % 2850 52.6 %
12 ComStock 1.0.2b 2720 74 75 52 43.3 % 2767 40.4 %
13 Komodo 2.03 2717 85 88 37 35.1 % 2823 43.2 %

If anyone wants to test these settings, here they are -

// Fire 2.2 Blaze fire.cfg

// System
Verbose_UCI 0
Write_Log 0
Hash 128
Pawn_Hash 32
Max_Threads 64
MultiPV 1
Random_Range 20

// RobboBases
Use_RobboBases 0
Verbose_RobboBases 0

// TotalBases
AutoLoad_TotalBases 0
TotalBase_Path c:\TotalBases
TotalBase_Cache 1

// TripleBases
AutoLoad_TripleBases 0
TripleBase_Path c:\TripleBases
TripleBase_Hash 1
Dynamic_TripleBase_Cache 1

// Eval Weights
Draw_Weight 100
King_Safety_Weight 131
Material_Weight 100
Mobility_Weight 115
Pawn_Weight 100
Positional_Weight 121
PST_Weight 100

// Lazy Eval
Lazy_Eval_Min 150
Lazy_Eval_Max 300

// Piece Values
Pawn_Value 100
Knight_Value 313
Bishop_Value 320
Rook_Value 512
Queen_Value 967
Bishop_Pair_Value 50

// Prune Thresholds
Prune_Check 10
Prune_Pawn 160
Prune_Minor 500
Prune_Rook 800

// Search Parameters
Aspiration_Window 8
Delta_Cutoff 25000
DepthRed_Min 12
Extend_In_Check 0
Height_Multiplier 64
History_Threshold 50
Low_Depth_Margin 1125
Min_Depth_Multiplier 48
Min_Trans_Move_Depth 16
Null_Reduction 9
QS_Alpha_Threshold 200
Search_Depth_Min 22
Search_Depth_Reduction 6
Top_Min_Depth 14
Undo_Count_Threshold 15
Value_Cut 15000
Verify_Null 1
Verify_Reduction 1

// Split Depths
Split_At_CN 1
AN_Split_Depth 14
CN_Split_Depth 16
PV_Split_Depth 12

// Time Management
Absolute_Factor 25
Battle_Factor 95
Easy_Factor 15
Easy_Factor_Ponder 30
Normal_Factor 80

// UCI Info strings
CPULoad_Info 0
CurrentMove_Info 0
Depth_Info 0
HashFull_Info 0
LowDepth_PVs 0
NPS_Info 0
TBHit_Info 0

32 or 64 bit?