IvanHoe V47 tests

from a scumortal tester



Games Completed = 500 (Avg game length = 18.542 sec)
Settings = RR/64MB/2000ms+150ms/M 2147483647cp for 5 moves, D 201 moves/EPD:d:\Progs\LittleBlitzer\openings_100.epd
Time = 3211 sec elapsed


1. Ivanhoe B47cAx64 236.5/500 126-153-221 (L: m=153 t=0 i=0 a=0) (D: r=170 i=25 f=20 s=4 a=2) (tpm=160.4 d=14.2 nps=2135370)

2. Ivanhoe-B 55mPPsse4.2 263.5/500 153-126-221 (L: m=126 t=0 i=0 a=0) (D: r=170 i=25 f=20 s=4 a=2) (tpm=158.2 d=14.5 nps=2132512)

~~~~~~~#######

Shows NO improvement over V55

Tests conducted on 2 quad pcs. ane AMD , other Intell
one with same books . other No books
Triplebases (OLD) for Ivan No base for houdini , ponder OFF
same books
[img][You must be registered and logged in to see this link.][/img]


No books
[img][You must be registered and logged in to see this link.][/img]

Sandi
International Master Elo 2352.


Intel QX 9775 Yorkfield (8X4090 MHz)
Time: 3'+0". Hash 256. Ponder off. GUI Fritz 12. Tablebases 3-4-5 + Robbo Triple Bases. Book: HS-7moves plus.ctg.

Deep Rybka 4 - IvanHoe B47cA X64 :
51/99 - 48/99 . 51.52% - 48.48% ( + 19 - 16 = 64 ). one crash.


@@@@@@@@@@@@@@

another backwards step

[img][You must be registered and logged in to see this link.][/img]


No books ,No Bases ,No Large pages , ponder Off

[img][You must be registered and logged in to see this link.][/img]

Its common knowledge that rykba is far stronger at longer time control games and there are reports of it ALLMOST matching Houdini1.5 at them on 12 core machines and clusters
10 minutes +1 second increment games can be very long games
these 13 took some 8 hours on our 3.4 ghz quad , thats more than enough wasted cpu time for us as it only shows what other tests do
ie B47cA X64 can match rykba4 but not beat it , and cannot beat houdini in tests either , its slow in comparision and without bases loses even more games

As KLOs version beat B47cA X64 quite covincingly it may well match houdini if its tested with TripleBases bases but its not reading them so we cant test that possibility

test was
NO books Triple bases for Ivan , NO Bases for Houdini , Large Page for BOTH , Ponder OFF

[img][You must be registered and logged in to see this link.][/img]

Haveing seen another test between these two engines we thought to test that tests results

test conditions No books No Large Pages No Bases just engines and NO HYPE

perhaps others pc had been ovedoing the vodka

"B" wrote:

[img][You must be registered and logged in to see this link.][/img]

no books .no bases , no good

K.L.0 maker of compile wrote..
IvanHoe 0B.01.08 (Windows x64 only) is a mod based on IvanHoe beta
999947c source code. It has an interesting dynamic style of play.
There's no reason to believe it is stronger, and as a matter of fact,
you might be better off not downloading it.


#########

we agree , Dont Bother even down loading it

[img][You must be registered and logged in to see this link.][/img]

09 version
K.L.O writes " with changed time management "

we say more of same, " slow and weak "

AMD 965 3.4Ghz , 8 gb memory
windows visa home prieum 64 bit
Fritz Bench Mark
Relative Speed 17.20
Kilo Nodes per second 8256
Gui Deep Fritz 12
Power Book 2011 (up to 10 moves )
Engine hash 256
Hiudini NO bases
Ivan WITH TripleBases
ponder OFF
5 minute Blitz





[img]

[You must be registered and logged in to see this link.][/img]

Houdini won this game , makeing it 6 wins

1 Ivanhoe B47cD +34 +42/=173/-19 54.91% 128.5/234
2 IvanHoe 47c GH x64 -34 +19/=173/-42 45.09% 105.5/234


1 minute time control,hash 128 megs,fritz 12 gui,large pages OFF,NO
bases,perfect 2010 up to move 7,default settings for both engines


games
new link [You must be registered and logged in to see this link.]