Crystal 7 CMB

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

Moderator: Ras

Jouni
Posts: 3656
Joined: Wed Mar 08, 2006 8:15 pm
Full name: Jouni Uski

Re: Crystal 7 CMB

Post by Jouni »

Crystal 7? (name is Crystal dev) has bigger net and it's speed about 50% of Crystal 6. No wonder version version 7 is clearly worse in all test suites. I go back to version 6 :) .
Jouni
jhellis3
Posts: 548
Joined: Sat Aug 17, 2013 12:36 am

Re: Crystal 7 CMB

Post by jhellis3 »

I think I'll quit this forum.
Jouni
Posts: 3656
Joined: Wed Mar 08, 2006 8:15 pm
Full name: Jouni Uski

Re: Crystal 7 CMB

Post by Jouni »

One example. Arasan suite 15s and 6 cores. Crystal 6 solved 193/200 with time 4m29s. Crystal 7 solved 185 with time 6m42s.
Jouni
User avatar
Eelco de Groot
Posts: 4671
Joined: Sun Mar 12, 2006 2:40 am
Full name:   Eelco de Groot

Re: Crystal 7 CMB

Post by Eelco de Groot »

jhellis3 wrote: Sat Nov 11, 2023 6:23 pm I think I'll quit this forum.
Sometimes I get depressed seeing what computerchess is doing. Not just here of course.
If there is more content, there tends to be less noise. That is just a natural reaction. Always been like that. Please do not give up on Crystal! It really was a highlight and lighting up the forum to read about and use! Thanks Joseph for the new version!
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
peter
Posts: 3410
Joined: Sat Feb 16, 2008 7:38 am
Full name: Peter Martan

Re: Crystal 7 CMB

Post by peter »

Jouni wrote: Sat Nov 11, 2023 7:41 pm One example. Arasan suite 15s and 6 cores. Crystal 6 solved 193/200 with time 4m29s. Crystal 7 solved 185 with time 6m42s.
Arasan- suite isn't difficult enough for engines like Crystal SMP on strong hardware with 15"/pos.
256 out of my old harder 128- suite and 128 from Dann's new one "pure.epd" with 15"/pos. and 8 threads (see end of Dann's thread about it way back then):

Code: Select all

Crystal 6 PMT

Bisher gelöst: 183 von 256  ;  27:55m

         1   2   3   4   5   6   7   8   9  10  11  12  13  14  15  16  17  18  19  20
 -------------------------------------------------------------------------------------
   0 |   1   -   -   8   8  12   0   -   8   -   0   3   0   0   0   8   -   3   -   8
  20 |   1   6   6   -   4   0   9   -   0   7   1  11   3   -   3   -   -   -   -   5
  40 |   -   3   0   0   -   -   8   -   -   -   -   7   3   -   -   -  15   1   4   0
  60 |   4   0   -   -   -   -   -   -   3   0   0   3   0   3   0   -   0   5   0   0
  80 |   3   0   0   1   3   2   3   0   -   0   2  10   0  15   -   0   -   -   1   1
 100 |   -   -   -   0   -   -   0   3   1   0   -   -   2   0   -   -   -  11   -   -
 120 |   -   -   -   -   -   -  14   0   -   0  15   0   0   0   1   0   -   0   0   0
 140 |   7   1   0   0   0   -   0   0   0   2   6   0   0   7   -   0   0   3   0   -
 160 |   3   -   9  14   0   1   1   -   3   0   3   7   0   0   -   -   2   1   0   7
 180 |   0   -   -   0   0   4  15   0   0   -   0   0  15  13   -   4   -   9   0   2
 200 |   0   -  10  11   -   2   5   3   0   -   0   0   3   0  13   0  15  15   0   4
 220 |   6   -   1   2   4   0   0   0   0   -   0   1   2   0   -   1   0   1   2   0
 240 |   0   2   6   0   2   0   0   0   0   -   1   0   0   0   0   0

  K/s: 10.546.361  
  TotTime: 38:22m    SolTime: 27:55m

Code: Select all

Crystal dev (7 CMB)

Bisher gelöst: 184 von 256  ;  24:11m


         1   2   3   4   5   6   7   8   9  10  11  12  13  14  15  16  17  18  19  20
 -------------------------------------------------------------------------------------
   0 |   0   -   -   1   2   0   0   -   0   -   0   1   0   0   0   -   -   4   -   -
  20 |   2  13   6   -  13   0   -   -   0   1   0   7   5   -   5   -  11   -   -   0
  40 |   -   -   0   0   -   -   3   -   -   0   -   3   7   5   -   -   -   0   6   0
  60 |   3   0   -   -   0   -   0   -   6   0   0   3   0   0   0   -   1   3   0   0
  80 |   1   0   0   0   0   3   3   0   4   0   2   1   1   -   -   0   -   -   8   0
 100 |   -   -   -   0   -   -   0   2   0   0   -   -   0   0   -   8   -   2   -   -
 120 |   -   -   -   -   -  10   6   0   -   0  11   -   0   2   3   0   -   0   0   0
 140 |   0   0   0   0   0   -   0   0   1   0   2   0   0   3   -   0   0   4   0   -
 160 |   3   -   0   4   0   0   -  10  12   -   2   4   0   0   -   0   1   0   3   0
 180 |   0   -   -   0   0   0   3   0   0   -   1   8   4   -   -   2   -   3   0   0
 200 |   0   -   2   0   -   0   3   9   0   0   0   0   -   3   8   0   1   0   0   0
 220 |   0   -   0   1   2   0   0   0   0   3   1   7   0   0   -   1   0   0   0   3
 240 |   0   5   3   0   0   0   0   0   0   -   0   0   0   0   0   0


  K/s: 11.230.389  
  TotTime: 33:58m    SolTime: 24:11m
WDL with faster or one-sided solved ones = wins, one-sided not or slowlier solved ones = losses, rest draws, counted manually like this:

Image

, and the numbers given into EolStat from the point of view of Crystal 7:

Code: Select all

Wins   = 76
Draws  = 142
Losses = 38
Av.Op. Elo = 3500

Result     : 147.0/256 (+76,=142,-38)
Perf.      : 57.4 %
Margins    :
68 %       : (+  2.0,-  2.0 %) -> [ 55.4, 59.4 %]
95 %       : (+  4.0,-  4.0 %) -> [ 53.4, 61.4 %]
99.7 %     : (+  6.0,-  6.1 %) -> [ 51.3, 63.4 %]

Elo        : 3552
Margins    :
68 %       : (+ 14,- 14) -> [3538,3566]
95 %       : (+ 29,- 28) -> [3524,3580]
99.7 %     : (+ 43,- 43) -> [3509,3595]
And a little match with UHO 6mvs. 110-119cp, single thread, 3'+1":

Score of Crystal 7 CMB vs Crystal 6 PMT: 116 - 97 - 287 [0.519]
Elo difference: 13.2 +/- 19.9, LOS: 90.4 %, DrawRatio: 57.4 %
500 of 500 games finished.
Peter.
Jouni
Posts: 3656
Joined: Wed Mar 08, 2006 8:15 pm
Full name: Jouni Uski

Re: Crystal 7 CMB

Post by Jouni »

Also in difficult HTC suite and longer time control Crystal 6 was better and faster.OK I need probably faster CPU.
Jouni