IvanHoeV57 /V55 Tests

First topic message reminder :

[img][You must be registered and logged in to see this link.][/img]
same books , ponder off
Whilst test is an unfair one as DR4 had no Namilov nor large pages we let it run on to test integrity of Ivan 57 and it was exellent, could be faster but , it was stable , no crashes and read bases
Anyoe who can enable Large page and or Namilov please test same engines

Yes it is very sad, we take no joy in posting this tests results

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

:gal4:

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

Another short and unfair test dr4 at a disadvantage as T63D had triple bases

One cannot say "no improvement " since V63's , as the addition of large pages options will benifit those that can acsess the permissions of there OS to have them

We realy need input from intell users but it seems they no longer care to contribute to testings , another reason for us to reconsider not closeing down this forum

Intel QX 9775 Yorkfield (8x4090 MHz).
Fritz Benchmark: relative speed = 38,58 . kn/s = 18.387.
Time : 3'+0". Hash 128. GUI Fritz 12 . Tablebases 5 pieces. Ponder off.
book: HS-7moves.ctg.

IvanHoe 57aU-x64 - Deep Rybka 4 : +11 -10 = 29.
A very good result.
Vlad0's 57aU is to 8 cores much stronger than 55mU


Sandi
International Master FIDE Elo 2352





:yo:

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


Fell asleep so a longer test , same books both with triplebases

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

This B57d_07 is an Intell compile made for AMD pc's NOT the one posted earlier

An example of how many games are needed perhaps as after 50 games i was sure these compiles were equally strong though the intell compiled one is slightly faster

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


this is peter pants recompileing version many are raveing about

no bases , same books . ponder off

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

the above engine was /is vlad0s intell compile of whmorlyjnrs engine and thanks to vlado0 it was briefly the #1 ivanhoe

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

older and popcount Ivan 54's have same names
[img][You must be registered and logged in to see this link.][/img]


Just three of our tests for Vlad0 when he asked us to test his intell compiled version of v57


this is probably the #1 ivanhoe at present KLOs latest is about on a par with it ...perhaps

Intel Xeon X5355 Octa (8x2660 MHz)
Time: 3'+0". Hash 128. Ponder off. GUI Fritz 12. Tablebases 5 pieces + Robbo TripleBases. book: HS-7moves.ctg.

IvanHoe 57aUI-x64 - Deep Rybka 4 : 25.5 - 24.5 .51% - 49%
(+10 -9 =31)

Good,solid work Vlad0 !

Sandi

###


AMD-X6.x64 Blitz:10', Blitz:10' 0



1 IvanHoe 57aUI-x64 3195 +2/-1/=17 52.50% 10.5/20
2 Deep Rybka 4 x64 3185 +1/-2/=17 47.50% 9.5/20

Skulltrail Intel QX 9775 Yorkfield (8x4090 MHz)
Time 3'+0". Hash 128. GUI Fritz 12. Ponder off. Tablebases 5 pieces + Robbo TripleBases.book:HS-7moves.ctg.

IvanHoe T55A(SSE3)x64 - Deep Rybka 4 :
25 - 25 .50%-50%. (+7 -7 = 36)

Sandi

~~

"B" wrote:

Skulltrail Intel QX 9775 Yorkfield (8x4090 MHz)
Time 3'+0". Hash 128. GUI Fritz 12. Ponder off. Tablebases 5 pieces + Robbo TripleBases.book:HS-7moves.ctg.

IvanHoe T55A(SSE3)x64 - Deep Rybka 4 :
25 - 25 .50%-50%. (+7 -7 = 36)

Sandi

~~

Intel Xeon Octa (8x2660 MHz)
Time 3'+0". Hash 128. GUI Fritz 12. Ponder off.Tablebases 5 pieces + Robbo TripleBases. Book:HS-7moves.ctg.

IVANHOE T55A (NO SSE)x64 - DEEP RYBKA 4 : 30 - 20 !!!
60% -40% (+16 -6 =28) !!!!!!!!

THE ABSOLUTELY BEST RESULT OF A IVANHOE COMPILE AGAINST
DEEP RYBKA 4 TO 8 THREADS WITH 8 PHISICAL CORES.


Sandi
~~~

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

T55B seems very stable . loads bases perfectly well .but not as strong as T55A in this short test
Ponder OFF, 512 engine hash , same books , Fritz11 GUI (more reliable than F12 )
AMD phenom 3,4ghz winows vista 64
# score now +6 =16 -2

First to 10 wins when margin not close is good indication as to which engine is stronger

+10 =23 -4

However T55B is more stable and faster than T55A so why it is weaker only Ahmed may know
T55A still

Some absolute RUBBISH claims on immortal forum about the W H Mowely Jnr engines being stronger than all Ivans is a JOKE

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