In an effort to provide compiles that will work on many different systems,
I've released:
Robbolito 0.21Q
distance-to-mate display fixed
5 fast Intel PGO compiles included:
RobboLito 0.21Q x64 SSE4.2.exe
RobboLito 0.21Q x64 SSE4.2 NLP.exe
RobboLito 0.21Q x64 NLP.exe
RobboLito 0.21Q w32.exe
RobboLito 0.21Q w32 NLP.exe
for a variety of systems
2-4% speed increase for most
please see the document 'compiles.rtf' for more info on the different choices
games:
mediafire.com download.php?a7p2j4a6y3w1an7
+ 45 ELO
www.chesslogik.com
if you're in the middle of testing, i recommend you change versions, 0.21Q is ~2-3% faster and has finished testing +7 stronger against IvanHoe than 0.20Q[url=games]
I've released:
Robbolito 0.21Q
distance-to-mate display fixed
5 fast Intel PGO compiles included:
RobboLito 0.21Q x64 SSE4.2.exe
RobboLito 0.21Q x64 SSE4.2 NLP.exe
RobboLito 0.21Q x64 NLP.exe
RobboLito 0.21Q w32.exe
RobboLito 0.21Q w32 NLP.exe
for a variety of systems
2-4% speed increase for most
please see the document 'compiles.rtf' for more info on the different choices
- Code:
Games Completed = 10229 of 50000 (Avg game length = 13.066 sec)
Settings = RR/32MB/1000ms+100ms/M 1000cp for 12 moves, D 150 moves/EPD:openings.epd(4000)
Time = 24595 sec elapsed, 95625 sec remaining
1. IvanHoe 9.46b x64 4459.0/10229 1486-2797-5946 (L: m=1474 t=8 i=0 a=1315) (D: r=4217 i=860 f=554 s=15 a=300) (tpm=98.0 d=13.14 nps=2960148)
2. RobboLito 0.21Q x64 SSE4.2 5770.0/10229 2797-1486-5946 (L: m=1048 t=2 i=0 a=436) (D: r=4217 i=860 f=554 s=15 a=300) (tpm=95.5 d=13.50 nps=3745487)
games:
mediafire.com download.php?a7p2j4a6y3w1an7
+ 45 ELO
- Code:
Wins = 2797
Draws = 5946
Losses = 1486
Av.Op. Elo = 3200
Result : 5770.0/10229 (+2797,=5946,-1486)
Perf. : 56.4 %
Margins :
68 % : (+ 0.3,- 0.3 %) -> [ 56.1, 56.7 %]
95 % : (+ 0.6,- 0.6 %) -> [ 55.8, 57.0 %]
99.7 % : (+ 0.9,- 0.9 %) -> [ 55.5, 57.3 %]
Elo : 3245
Margins :
68 % : (+ 2,- 2) -> [3243,3247]
95 % : (+ 4,- 4) -> [3240,3249]
99.7 % : (+ 7,- 7) -> [3238,3251]
www.chesslogik.com
if you're in the middle of testing, i recommend you change versions, 0.21Q is ~2-3% faster and has finished testing +7 stronger against IvanHoe than 0.20Q[url=games]
Last edited by kranium on Tue Oct 02, 2012 11:34 am; edited 7 times in total
Had to start the tournament over issues with the computer, same position start from 1. d4 Nf6 2.c4 e5 3.dxe5 Ne4, after the defrag seems to be running smoothly now and I added a couple of engines so we shall see. I also shortened the tornament to two rounds. I think Critter will do much better now that it is not running at 1kn/sec.
. Robbo 2.1 looks very strong so far in my G/25 10. It is a theam tournament Fed - Budapest 20 engines it is in second behind IvanHoe 47C GH and delt Ivan it's solo loss as of 11 rounds. Suprisingly my engine that has performed really better than any up until now (Critter 1.6a) is not doing well at all 1 -4 the rest =. Houdini 1.5a 2-1 =7(after 10) not doing that well either has almost drawn all so far. In my tournaments though Critter usually has more wins anyway while Houdini draws alot rarely lossing though. I really expect Critter to do well though since most of the higher rated engines it has played leaving the bottom to go now.
