TESTING Ivanhoe-BetaWH_25

[Tomcass]

TESTING Ivanhoe-BetaWH_25

TEST 1

Quad 2.33 256 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 2.0
Time Control: 5 min+ 0 sec.
4 cores for both Ivans, 1 core for Gull12
80 games for IvanBetaWH_25, 120 games overall.


1 Ivanhoe-BetaWH_25_w32(x4) 24.0 - 16.0 33.5 - 6.5 57.5/80
2 IvanHoe T52B 16.0 - 24.0 33.0 - 7.0 49.0/80
3 Gull012 6.5 - 33.5 7.0 - 33.0 13.5/80


TEST 2.


I7 975 1,024 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 1.0
Time Control: 5 min+ 0 sec.
4 cores with 4 hyperthread for IvanBetaWH_25
4 cores NO hyperthread for IvanT52B
1 core for Gull12
80 games for IvanBetaWH_25, 120 games overall.


1 IvanHoe T52B(x4) 23.5 - 16.5 32.0 - 8.0 55.5/80
2 Ivanhoe-BetaWH_25_w64(X8) 16.5 - 23.5 34.5 - 5.5 51.0/80
3 Gull012w64 8.0 - 32.0 5.5 - 34.5 13.5/80


GLOBAL SCORE AFTER 160 GAMES FOR IvanHoeBetaWH_25, 240 GAMES OVERALL.

1 Ivanhoe-BetaWH_25 40.5 – 39.5 68.0 - 12.0 108.5/160 67.81%
2 IvanHoe T52B 39.5 - 40.5 65.0 - 15.0 104.5/160 65.31%
3 Gull012 12.0 - 68.0 15.0 – 65.0 27.0/160 16,88%

Regards,

Tom.

[Tomcass]

TESTING Ivanhoe-BetaWH_25 (2)

TEST 1

Quad 2.33 256 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 2
Time Control: 5 min+ 0 sec.
4 cores
80 games

2010-07Beta_25_Bill

Ivanhoe-BetaWH_25_w32(x4) - Deep Rybka 4 w32 23.5 - 16.5 +17/=13/-10 58.75%
Ivanhoe-BetaWH_25_w32(x4) - Houdini 1.03a w32 4_CPU 19.5 - 20.5 +10/=19/-11 48.75%

TEST 2.

I7 975 1,024 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 2
Time Control: 5 min+ 0 sec.
4 cores and 8 threads
80 games
201007IvanWMH25w64

Ivanhoe-BetaWH_25_w64(X8) - Deep Rybka 4 x64 19.0 - 21.0 +7/=24/-9 47.50%
Ivanhoe-BetaWH_25_w64(X8) - Houdini 1.03a x64 8_CPU 20.5 - 19.5 +4/=33/-3 51.25%

GLOBAL SCORE AFTER 160 GAMES.

Ivanhoe-BetaWH_25 - Deep Rybka 4 42.5 -37.5 +24/=37/-19 53.12%
Ivanhoe-BetaWH_25 - Houdini 1.03a 40.0 - 40.0 +14/=52/-14 50.00%

Regards,

Tom.

[jpqy]

Hi Tom,

Always great to follow your testings..Thanks!

JP.

[gerold]

Thanks for the tests Tom.

In my 5/3 TC with 200 games Ivanhoe T52A is = to Houdini 1.3.

Looks like there are 2 engines about 15-20 elo stronger than
R4.

Best,
Gerold.

[Dr.Wael Deeb]

Thanks for the tests Tom.

In my 5/3 TC with 200 games Ivanhoe T52A is = to Houdini 1.3.

Looks like there are 2 engines about 15-20 elo stronger than
R4.

Best,
Gerold.

I can speculate that they are more than 2 engines Gerold and by that stronger than Houdini itself
Stay tunned regards,
Dr.D

[Tomcass]

Thanks for your comment, Jean Paul!.

TESTING Ivanhoe-BetaWH_24

TEST 1

Quad 2.33 256 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 2.0
Time Control: 5 min+ 0 sec.
4 cores for every engine
120 games

2010-08Beta_24Gauntlet(5min)-2 2010

Ivanhoe-BetaWH_24_w32 - Deep Rybka 4 w32 22.0 - 18.0 +13/=18/-9 55.00%
Ivanhoe-BetaWH_24_w32 - Houdini 1.03a w32 18.0 - 22.0 +8/=20/-12 45.00%
Ivanhoe-BetaWH_24_w32 - IvanHoe T52B 21.0 - 19.0 +6/=30/-4 52.50%

TEST 2.

I7 975 1,024 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 1.0
Time Control: 5 min+ 0 sec.
Hyper Thread: Off 4 cores each engine.
120 games

201008IvanWMH24w64_5'0Gauntlet 2010

Ivanhoe-BetaWH_24_w64(x4) - Houdini 1.03a x64 18.0 - 22.0 +4/=28/-8 45.00%
Ivanhoe-BetaWH_24_w64(x4) - IvanHoe T52B(x4) 19.5 - 20.5 +4/=31/-5 48.75%
Ivanhoe-BetaWH_24_w64(x4) - Deep Rybka 4 x64 (x4) 21.5 - 18.5 +9/=25/-6 53.75%


GLOBAL SCORE AFTER 240 GAMES FOR IvanBetaWH_24: 50,00%. 120/240.

Ivanhoe-BetaWH_24 - Deep Rybka 4 43,5 – 36,5 +22/=33/-15 54.37%
Ivanhoe-BetaWH_24 - Houdini 1.03a 36.0 – 44.0 +12/=48/-20 45.00%
Ivanhoe-BetaWH_24 - IvanHoe T52B 40.5 – 39.5 +10/=61/-9 50.63%

By now, the best version of WH Betas, despite the hard loss against Houdini 1.03a.

Currently testing WH_29, two thirds of test completed. Still better results than WH 24. Great work Bill !!

Regards,

Tom.

[Tomcass]

TESTING Ivanhoe-BetaWH_29

TEST 1

Quad 2.33 256 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 2.0
Time Control: 5 min+ 0 sec.
120 games
2010-08Beta_29Gauntlet(5min) 2010

Ivanhoe-BetaWH_29_w32 - Deep Rybka 4 w32 20.5 - 19.5 +10/=21/-9 51.25%
Ivanhoe-BetaWH_29_w32 - Houdini 1.03a w32 18.5 - 21.5 +6/=25/-9 46.25%
Ivanhoe-BetaWH_29_w32 - IvanHoe T52B 21.0 - 19.0 +8/=26/-6 52.50%

TEST 2.

I7 975 1,024 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 1.0
Time Control: 5 min+ 0 sec.
180 games
Ivanhoe-BetaWH_29_w64(x4) - Deep Rybka 4 x64 (x4) 32.0 - 28.0 +13/=38/-9
Ivanhoe-BetaWH_29_w64(x4) - Houdini 1.03a x64 (x4 ) 28.5 -31.5 +10/=37/-13
Ivanhoe-BetaWH_29_w64(x4) - IvanHoe T52B(x4) 28.0 - 32.0 +5/=46/-9


GLOBAL SCORE AFTER 300 GAMES FOR IvanBetaWH_29: 49.50% 148.5 /300.

Ivanhoe-BetaWH_24 - Deep Rybka 4 52.5 - 47.5
Ivanhoe-BetaWH_24 - Houdini 1.03a 47.0 - 53.0
Ivanhoe-BetaWH_24 - IvanHoe T52B 49.0 - 51.0

Regards,

Tom.

[Tomcass]

Wooops, a typo in my previous post. Sorry!:

The global score is obviously for WH_29 rather than 24. That is:

GLOBAL SCORE AFTER 300 GAMES FOR IvanBetaWH_29: 49.50% 148.5 /300.

Ivanhoe-BetaWH_29 - Deep Rybka 4 52.5 - 47.5
Ivanhoe-BetaWH_29 - Houdini 1.03a 47.0 - 53.0
Ivanhoe-BetaWH_29 - IvanHoe T52B 49.0 - 51.0

Regards,

Tom.

[Tomcass]

TESTING Ivanhoe-BetaWH_32

TEST 1

Quad 2.33 256 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 2.0
Time Control: 5 min+ 0 sec.

120 games

Ivanhoe-BetaWH_32_w32(x4) - Deep Rybka 4 w32 17.0 - 23.0 +10/=14/-16 42.50%
Ivanhoe-BetaWH_32_w32(x4) - Houdini 1.03a w32 19.0 - 21.0 +9/=20/-11 47.50%
Ivanhoe-BetaWH_32_w32(x4) - IvanHoe T52B 20.5 - 19.5 +9/=23/-8 51.25%

TEST 2.

I7 975 1,024 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 1.0
Time Control: 5 min+ 0 sec.

120 games

Ivanhoe-BetaWH_32_w64 - Deep Rybka 4 x64 (x4) 23.0 - 17.0 +11/=24/-5 57.50%
Ivanhoe-BetaWH_32_w64 - Houdini 1.03a x64 8_CPU (x4) 20.0 - 20.0 +5/=30/-5 50,00%
Ivanhoe-BetaWH_32_w64 - IvanHoe T52B(x4) 21.0 - 19.0 +4/=34/-2 52.50%


GLOBAL SCORE AFTER 240 GAMES FOR IvanBetaWH_32: 50.21% 120.5 /240

Ivanhoe-BetaWH_32 - Deep Rybka 40.0 – 40.0
Ivanhoe-BetaWH_32 - Houdini 1.03a 39.0 – 41.0
Ivanhoe-BetaWH_32 - IvanHoe T52B 41.5 – 38.5

After a very good start in the first third of this test, I expected a better performance for this version. Anyway, getting a bit more than 50% against these three powerful rivals gives me a positive feeling about the future development of these WH compiles.

Regards,

Tom.

[Tomcass]

TESTING Ivanhoe-BetaWH_35

TEST 1

Quad 2.33 256 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 2.0
Time Control: 5 min+ 0 sec.
120 games

Ivanhoe-BetaWH_35_w32(x4) - Deep Rybka 4 w32 22.5 - 17.5 +14/=17/-9 56.25%
Ivanhoe-BetaWH_35_w32(x4) - Houdini 1.03a w32 18.5 - 21.5 +7/=23/-10 46.25%
Ivanhoe-BetaWH_35_w32(x4) - IvanHoe T52C(x4) 23.0 - 17.0 +10/=26/-4 57.50%

(Included previous test of WH_33 as a reference for comparison)
Ivanhoe-BetaWH_33_w32(x4) - Deep Rybka 4 w32 21.0 - 19.0 +11/=20/-9 52.50%
Ivanhoe-BetaWH_33_w32(x4) - Houdini 1.03a w32 18.0 - 22.0 +6/=24/-10 45.00%
Ivanhoe-BetaWH_33_w32(x4) - IvanHoe T52C(x4) 20.0 - 20.0 +7/=26/-7 50.00%

TEST 2.

I7 975 1,024 hash
Gui: Fritz 12
Ponder: Off
No Triple or Robbo Bases
Book: HS Masterbook 1.0
Time Control: 5 min+ 0 sec.
120 games

Ivanhoe-BetaWH_35_w64 - Deep Rybka 4 x64 (x4) 19.0 - 21.0 +5/=28/-7 47.50%
Ivanhoe-BetaWH_35_w64 - Houdini 1.03a x64 (x4) 20.0 - 20.0 +6/=28/-6 50.00%
Ivanhoe-BetaWH_35_w64 - IvanHoe T52C(x4) 22.0 - 18.0 +6/=32/-2 55.00%


Ivanhoe-BetaWH_33_w64_SSSE4.2 - Deep Rybka 4 x64 18.5 - 21.5 +3/=31/-6 46.25%
Ivanhoe-BetaWH_33_w64_SSSE4.2 - Houdini 1.03a x64 20.0 - 20.0 +5/=30/-5 50.00%
Ivanhoe-BetaWH_33_w64_SSSE4.2 - IvanHoe T52C(x4) 18.0 - 22.0 +2/=32/-6 45.00%

GLOBAL SCORE AFTER 240 GAMES FOR IvanBetaWH_35: 52.08% 125.0 /240.

Ivanhoe-BetaWH_35 - Deep Rybka 41.5 – 38.5 51.87%
Ivanhoe-BetaWH_35 - Houdini 1.03a 38.5 – 41.5 48.12%
Ivanhoe-BetaWH_35 - IvanHoe T52C 45.0 – 35.0 56.25%

Ivanhoe-BetaWH_33 - Deep Rybka 39,5 – 40.5 49.37%
Ivanhoe-BetaWH_33 - Houdini 1.03a 38.0 – 42.0 47.50%
Ivanhoe-BetaWH_33 - IvanHoe T52C 38.0 – 42.0 47.50%

I know that 240 games are not enough to be sure about the strength of an engine. But BetaWH_35, still a bit below Houdini1.03a at this time control, is the best Ivan performer I have ever found at any of my tests. The difference over the very strong Ahmed’s T52C seems very relevant to me.

Regards,

Tom.