GM and Rybka vs. Stockfish

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

Moderators: hgm, Rebel, chrisw

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 11:58 am
Location: Antalya/Turkey

Re: GM and Rybka vs. Stockfish

Post by Sedat Canbaz »

Uri Blass wrote:
The game between GM and Rybka vs stockfish was not blitz so when I said that I believe top engines are at level of 3200 I did not mean to blitz conditions of 3 minutes per game but to the time control that the GM played in the match when he and rybka lost 3.5:0.5 that is practically 45 minutes per game+30 seconds per move.
Yes...but I say again we have no any info about the hardware conditions (annotations, depth ,time etc...), right ?

I mean Stockfish and Rybka on what kind of hardwares are played ?
I wonder a lot too about both used hardwares were identical or not ?!

I think it is not required to be mention again that depending on the processor speed any engine can be performed between 0-1000 Elo or even more...

Otherwise...without this info...I don't care too much about the current played results 3.5:0.5
Last edited by Sedat Canbaz on Fri Aug 15, 2014 11:12 am, edited 1 time in total.
Uri Blass
Posts: 10282
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: GM and Rybka vs. Stockfish

Post by Uri Blass »

Sedat Canbaz wrote:
Uri Blass wrote:
The game between GM and Rybka vs stockfish was not blitz so when I said that I believe top engines are at level of 3200 I did not mean to blitz conditions of 3 minutes per game but to the time control that the GM played in the match when he and rybka lost 3.5:0.5 that is practically 45 minutes per game+30 seconds per move.
Yes...but I say again we have no any info about the hardware conditions, right ?

I mean Stockfish and Rybka on what kind of hardwares are played ?
I wonder a lot too about both used hardwares were identical or not ?!

I think it is not required to be mention again that depending on the processor speed any engine can be performed between 0-1000 Elo or even more...

Otherwise...without this info...I don't care too much about the current played results 3.5:0.5
We have information about the hardware conditions

From the last post at page 3 of this thread:

"The 8-core Mac Pro would be somewhere around 16x as powerful as the MacBook (Rybka 3 ran on one core). That's a rough estimate based on the Mac Pro having 8x as many active cores, and ~2x the performance per core (Ivy Bridge Xeon @ 3Ghz vs Core2 @ 2Ghz). "
Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 11:58 am
Location: Antalya/Turkey

Re: GM and Rybka vs. Stockfish

Post by Sedat Canbaz »

Uri Blass wrote:
Sedat Canbaz wrote:
Uri Blass wrote:
The game between GM and Rybka vs stockfish was not blitz so when I said that I believe top engines are at level of 3200 I did not mean to blitz conditions of 3 minutes per game but to the time control that the GM played in the match when he and rybka lost 3.5:0.5 that is practically 45 minutes per game+30 seconds per move.
Yes...but I say again we have no any info about the hardware conditions (annotations, depth ,time etc...), right ?

I mean Stockfish and Rybka on what kind of hardwares are played ?
I wonder a lot too about both used hardwares were identical or not ?!

I think it is not required to be mention again that depending on the processor speed any engine can be performed between 0-1000 Elo or even more...

Otherwise...without this info...I don't care too much about the current played results 3.5:0.5
We have information about the hardware conditions

From the last post at page 3 of this thread:

"The 8-core Mac Pro would be somewhere around 16x as powerful as the MacBook (Rybka 3 ran on one core). That's a rough estimate based on the Mac Pro having 8x as many active cores, and ~2x the performance per core (Ivy Bridge Xeon @ 3Ghz vs Core2 @ 2Ghz). "

Oh...yes right now I noticed this info...thanks

But however this would be completely wrong view,
If we look at current results (3.5: 0.5), only 4 games, plus Rybka 3 is played with human

And I don't think that the current duel (Rybka + Human vs Stockfish)
is a good indicator about determining Stockfish should be rated around 3200 Elo at slow time control

I say again the solution is Challenge )))
Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 11:58 am
Location: Antalya/Turkey

Re: GM and Rybka vs. Stockfish

Post by Sedat Canbaz »

Btw, still I could not find any info for the current played games to be with annotations, depth ,time etc...

Why a such data is missing ??
Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 11:58 am
Location: Antalya/Turkey

Re: GM and Rybka vs. Stockfish

Post by Sedat Canbaz »

One thing more,
I agree that Stockfish probably will be rated around 3200 Elo, but on my AMD 3400 2.40 GHz ))

But what about Stockfish on i7 980X @4.33GHz ?? can we expect same performance ???

Btw, Stockfish 5 is rated 3287 Elo, in CCRL (40/40 AMD 4600 2.40GHz):
http://computerchess.org.uk/ccrl/4040/r ... t_all.html

In other words,
CCRL ratings confirm my point of view... !
Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 11:58 am
Location: Antalya/Turkey

Re: GM and Rybka vs. Stockfish

Post by Sedat Canbaz »

I hope the below bench results will be helpful to compare the processor speeds

Fritz Chess Benchmarks:

Code: Select all

Hardware-Processor        Speed      Cores     kN/s
Intel Core i7 980X      @ 4.33 GHz     6      18709
Intel Core 2 Q6600        2.40 GHz     4       6771
AMD Athlon 64 X2 4600+    2.40 GHz     2       2695
AMD Athlon 64 3400+       2.40 GHz     1       1367
For full list:
http://www.sedatcanbaz.com/chess/?page_id=287

After checking the above results,
I decided to calculate again the speed and the Elo performance

If we take as an example that doubling the processor speed is average 75 Elo

1350 kns = 3200 Elo
2700 kns = 3275 Elo
5400 kns = 3350 Elo
10800 kns = 3425 Elo
21600 kns = 3500 Elo


Note that we should not forget to mention again the power of the opening books
And with a very strong book, Stockfish would be rated at least +100 Elo more

In other words (as I mentioned before),
Stockfish is expecting to be rated above 3500 Elo (on decent fast hardware + superior strong book)


Hopes this helps too!)
User avatar
Laskos
Posts: 10948
Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos

Re: GM and Rybka vs. Stockfish

Post by Laskos »

Sedat Canbaz wrote:I hope the below bench results will be helpful to compare the processor speeds

Fritz Chess Benchmarks:

Code: Select all

Hardware-Processor        Speed      Cores     kN/s
Intel Core i7 980X      @ 4.33 GHz     6      18709
Intel Core 2 Q6600        2.40 GHz     4       6771
AMD Athlon 64 X2 4600+    2.40 GHz     2       2695
AMD Athlon 64 3400+       2.40 GHz     1       1367
For full list:
http://www.sedatcanbaz.com/chess/?page_id=287

After checking the above results,
I decided to calculate again the speed and the Elo performance

If we take as an example that doubling the processor speed is average 75 Elo

1350 Kns = 3200 Elo
2700 kns = 3275 Elo
5400 kns = 3350 Elo
10800 kns = 3425 Elo
21600 kns = 3500 Elo


Note that we should not forget to mention again the power of the opening books
And with a very strong book, Stockfish would be rated at least +100 Elo more

In other words (as I mentioned before),
Stockfish is expecting to rated above 3500 Elo (on decent fast hardware + superior strong book)


Hopes this helps too!)
With so many successive posts, are you not realizing that the issue was about FIDE rating of Stockfish, not computer chess ratings?
Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 11:58 am
Location: Antalya/Turkey

Re: GM and Rybka vs. Stockfish

Post by Sedat Canbaz »

Laskos wrote:
Sedat Canbaz wrote:I hope the below bench results will be helpful to compare the processor speeds

Fritz Chess Benchmarks:

Code: Select all

Hardware-Processor        Speed      Cores     kN/s
Intel Core i7 980X      @ 4.33 GHz     6      18709
Intel Core 2 Q6600        2.40 GHz     4       6771
AMD Athlon 64 X2 4600+    2.40 GHz     2       2695
AMD Athlon 64 3400+       2.40 GHz     1       1367
For full list:
http://www.sedatcanbaz.com/chess/?page_id=287

After checking the above results,
I decided to calculate again the speed and the Elo performance

If we take as an example that doubling the processor speed is average 75 Elo

1350 Kns = 3200 Elo
2700 kns = 3275 Elo
5400 kns = 3350 Elo
10800 kns = 3425 Elo
21600 kns = 3500 Elo


Note that we should not forget to mention again the power of the opening books
And with a very strong book, Stockfish would be rated at least +100 Elo more

In other words (as I mentioned before),
Stockfish is expecting to rated above 3500 Elo (on decent fast hardware + superior strong book)


Hopes this helps too!)
With so many successive posts, are you not realizing that the issue was about FIDE rating of Stockfish, not computer chess ratings?
Yes, sure.... I make many successive posts, because some people don't read or don't understand for what I am talking....

Yes....I am meaning especially about FIDE Ratings, e.g Stockfish should be rated around 3500 Elo on latest fast hardwares + super strong book (in FIDE rating)

Image
.
And I say again: CHALLENGE

.
Otherwise its pointless to discuss any more with people who does not agree with me, ok ?)
Uri Blass
Posts: 10282
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: GM and Rybka vs. Stockfish

Post by Uri Blass »

Note that fide rating is not about blitz of 3 minutes per game.

I think that no human can win in your challange of 3 minutes per game
but it does not mean that they cannot do it at longer time control.

90+30 is the normal time control that I play for fide rating.
elcabesa
Posts: 855
Joined: Sun May 23, 2010 1:32 pm

Re: GM and Rybka vs. Stockfish

Post by elcabesa »

instead of writing CHALLENGE everywhere you could start sending email to some of the top players and trying to get some sponsor. :)

it could be a very important match