4.7 GHz IBM Power6 dual-core processor

[tiger]

Wow!

Thank you Eelco.


// Christophe

[Tony]

Hallo,

CPU speed it´s MORE irrelevant THAT chess knowledge

You the right also the wrongs... you saying any many thinks? saying Formula-1 racing use the pentium-4 wins
the core-to-dua? Wrong you You saying you go war in battle ride donkey can live after battling Armour Tanker. You sure die and cannot see the homeland welcome

You wisdom many I see, but many here must buy decipher software to benefit your computer chess gems wisdom. First translate you english to understanding english then use you wisdom gems...
and elo straight shoot to heaven


I Snailchess beat Rybka 3 ways:-
1) buy very expensive hardware(1 million eros) very powerful and pinned down Vasik. Tell him sorry I rich you poor I win.
2) I poor use pentium4. Vasik use 8-core. I win and tell Vasik sorry. I win I very smart you lesser smart.
I add many knowlege in Snailchesss - NoNoNo! not speed probelm... smart problem.
3) ....

Better Regards,
Rasjid

I'm going to leave this thread and let you discuss together, you seem to understand each other quite well, but me has got eine grosse headache.


// Christophe

Me think yes too.

Tony

[Uri Blass]

Hallo,

CPU speed it´s MORE irrelevant THAT chess knowledge

You the right also the wrongs... you saying any many thinks? saying Formula-1 racing use the pentium-4 wins
the core-to-dua? Wrong you You saying you go war in battle ride donkey can live after battling Armour Tanker. You sure die and cannot see the homeland welcome

You wisdom many I see, but many here must buy decipher software to benefit your computer chess gems wisdom. First translate you english to understanding english then use you wisdom gems...
and elo straight shoot to heaven


I Snailchess beat Rybka 3 ways:-
1) buy very expensive hardware(1 million eros) very powerful and pinned down Vasik. Tell him sorry I rich you poor I win.
2) I poor use pentium4. Vasik use 8-core. I win and tell Vasik sorry. I win I very smart you lesser smart.
I add many knowlege in Snailchesss - NoNoNo! not speed probelm... smart problem.
3) ....

Better Regards,
Rasjid

I'm going to leave this thread and let you discuss together, you seem to understand each other quite well, but me has got eine grosse headache.


// Christophe

I agree.
I can only say that it seems that there is a misunderstanding and it seems that oliver understood that you said that speed is the only important thing in chess because you said that it is a speed problem.

It is clear that you did not say that nothing can be gained by better search or better evaluation.

It is also clear that speed is one of the things that is important.
Claiming that speed is less important than knowledge is meaningless
because it is impossible to compare.

The fact that rybka on 1ghz can beat Fritz5.32 on 4ghz is irrelevant because I do not see why taking Rybka and Fritz5.32 (you could also take rybka and rybkabeta that has less knowledge) and I do not see why taking ratio of 4:1(you could take speed ratio of 100:1)

The point is that cases when one side has speed advantage and another side has knowledge advantage are simply irrelevant because there is no rule how much knowledge advantage you give and there is no rule how much speed advantage you give.

Uri

[Karmazen & Oliver]

Yes, you know more than the top programmers

fantastic post... other POST in count... near 9.000 post...

in this theme you write a lot of post that not have ANY information. OK

it´s your problem... but since I write in this FORUM... I don´t see you anything interesting. or useful....

your intention is ALONE to cause. for that has not forgotten that the images are not LOADED in the talkches SERVER, in the bottom..

it is very sad that is you unable to argue anything in this sense and even so... so often write...

chess, be possibly a problem " knowledge + speed "... 70 % - 30 %

in your case ( human errors ).. if it is a problem of speed.

You WRITE but quick of what you thinks.

you possibly read a little MORe...

bye. since Spain.

[Karmazen & Oliver]

Hi, URI...

CPU speed it´s MORE irrelevant THAT chess knowledge

I agree.
I can only say that it seems that there is a misunderstanding and it seems that oliver understood that you said that speed is the only important thing in chess because you said that it is a speed problem.

It is clear that you did not say that nothing can be gained by better search or better evaluation.

It is also clear that speed is one of the things that is important.
Claiming that speed is less important than knowledge is meaningless
because it is impossible to compare.

The fact that rybka on 1ghz can beat Fritz5.32 on 4ghz is irrelevant because I do not see why taking Rybka and Fritz5.32 (you could also take rybka and rybkabeta that has less knowledge) and I do not see why taking ratio of 4:1(you could take speed ratio of 100:1)

The point is that cases when one side has speed advantage and another side has knowledge advantage are simply irrelevant because there is no rule how much knowledge advantage you give and there is no rule how much speed advantage you give.

Uri

I´m not agree...

because the theme TITLE Is:

" --- 4.7 GHz IBM Power6 dual-core processor --- "

If somebody write about this ... CPU, and I write that this increment of speed is not outstanding... that the chess is not a problem that is solved with THAT processor of IBM

and Cristopher... wirte that:

A computer chess expert is speaking ... 32-men tablebase ?¡ ... You do not really know what you are talking about ... 30 years back in the pas ...

OK... maybe you think that :

"...Claiming that speed is less important than knowledge is meaningless
because it is impossible to compare..."

but I don´t think that... it´s easy DO that test... with rybka 1. 0 or rybka 2.3.2n .... if you like more this example... rybka 2.3.2n have a lot of ELO, and NOT need 8 cores for that...

win other engines in 70-30 % of victorys it´s a lot of diferents to "only" speed problem...

you say that:

"...I do not see why taking ratio of 4:1(you could take speed ratio of 100:1)..."

I´m used this EXAMPLE... int ratio 4:1 .. 1 Ghz. of 4 GHz... ( not is lineal OK... but... it´s dificult find a CPU that you have change speed more... ) I can do that in my intel... I can change the BUS, and do a FAST CPU or slowly CPU...

The question is that I´m sued this example because the post TITLE is:

"...4.7 GHz IBM Power6 dual-core processor .."

and the autor of post question about if THIS CPU... it implied a great improvement in chess (he/she asks) ...

This CPU is X2... or X4 more fast that currents CPU... I need do a comparation similar... NOT is perfect OK ¡¡¡ but is BETtER that NOTHING ¡

My opinion is that NOT... THAT it is not the solution... for the time now ...

clear... that if you want to make tables it bases of 32 pieces... uff... pluf.

Ok... NOW... some programers think that chess knowledge in HIS engine are "complete"... ok... test that then whit a FAST CPU... and the other with a slowly CPU... or give more time to analysis...

I estimate... knowledge 70% speed 30%... (if it is not bigger even the difference... ) can that estimates can be made... but makes many many YEARS that ALL people decided that the gross force (speed) it WAS NOT the road for the chess...

Bye , since SPAIN.

[Karmazen & Oliver]

The only exception is when you can search until you get absolutely exact evaluations at the leaves of your search tree. In other words you need to have the X-men tablebases, X being the number of pieces left on the board, or an exact evaluator for your position. If all you have is the (X-1)-men tablebase, you still need to do a significant search job (and there is no guarantee in this case that all the leaves of your search tree will get exact evaluations).


We only have tablebases for positions with at most 6 pieces. That is, we have tablebases for small minority of chess positions. In all positions with 7 pieces left or more, search speed is crucial.

Based on this, my interpretation is that chess is definitely a speed problem, no matter how cleverly you search and how much positional knowledge you put in your program.
// Christophe

ONLY is correct near end-game... in the middle game... (after books) tacticals are more revelat ... and strategy ...

and ALL engines can use end-tables... not only rybka for example... SAY NOW that " win" the engine that fisrt come to "end-table" is a curious (interesting also) argument...

maybe Try THAT engine not "cames" near end with a bad positions or "structural weaknesses in the position".. If your afirmation was correct ... ALL GAMES finish in 0.5-0.5 ... and this not is TRUE... OK... something of good luck is necesary in select search...

bye.

PD: but... but I understand a little that that your you mean ( spanish-english is dificult for me )... I hope you are not right...

I find a lot of diferents style among the engines ... previus end games...

bye. and thanks for your time. but we are waiting about your post theme "ELO diferents" in evaluation incorrect engines... remember ? you said that in future you write your opinion about "the problems" in evaluation test engines.

[Terry McCracken]

Yes, you know more than the top programmers

fantastic post... other POST in count... near 9.000 post...

It's to the point Oliver...oh and soon I'll have written 10000 posts!

in this theme you write a lot of post that not have ANY information. OK

it´s your problem... but since I write in this FORUM... I don´t see you anything interesting. or useful....

Speak for yourself!

your intention is ALONE to cause. for that has not forgotten that the images are not LOADED in the talkches SERVER, in the bottom..

it is very sad that is you unable to argue anything in this sense and even so... so often write...

This has been gone over ad nauseam, I don't wish to go over it again and again.

chess, be possibly a problem " knowledge + speed "... 70 % - 30 %

in your case ( human errors ).. if it is a problem of speed.

It's both, end of story


You WRITE but quick of what you thinks.

you possibly read a little MORe...

bye. since Spain.

Yes I don't waste time on details unless I must.

[Uri Blass]

Hi, URI...

CPU speed it´s MORE irrelevant THAT chess knowledge

I agree.
I can only say that it seems that there is a misunderstanding and it seems that oliver understood that you said that speed is the only important thing in chess because you said that it is a speed problem.

It is clear that you did not say that nothing can be gained by better search or better evaluation.

It is also clear that speed is one of the things that is important.
Claiming that speed is less important than knowledge is meaningless
because it is impossible to compare.

The fact that rybka on 1ghz can beat Fritz5.32 on 4ghz is irrelevant because I do not see why taking Rybka and Fritz5.32 (you could also take rybka and rybkabeta that has less knowledge) and I do not see why taking ratio of 4:1(you could take speed ratio of 100:1)

The point is that cases when one side has speed advantage and another side has knowledge advantage are simply irrelevant because there is no rule how much knowledge advantage you give and there is no rule how much speed advantage you give.

Uri

I´m not agree...

because the theme TITLE Is:

" --- 4.7 GHz IBM Power6 dual-core processor --- "

If somebody write about this ... CPU, and I write that this increment of speed is not outstanding... that the chess is not a problem that is solved with THAT processor of IBM

and Cristopher... wirte that:

A computer chess expert is speaking ... 32-men tablebase ?¡ ... You do not really know what you are talking about ... 30 years back in the pas ...

OK... maybe you think that :

"...Claiming that speed is less important than knowledge is meaningless
because it is impossible to compare..."

but I don´t think that... it´s easy DO that test... with rybka 1. 0 or rybka 2.3.2n .... if you like more this example... rybka 2.3.2n have a lot of ELO, and NOT need 8 cores for that...

win other engines in 70-30 % of victorys it´s a lot of diferents to "only" speed problem...

you say that:

"...I do not see why taking ratio of 4:1(you could take speed ratio of 100:1)..."

I´m used this EXAMPLE... int ratio 4:1 .. 1 Ghz. of 4 GHz... ( not is lineal OK... but... it´s dificult find a CPU that you have change speed more... ) I can do that in my intel... I can change the BUS, and do a FAST CPU or slowly CPU...

The question is that I´m sued this example because the post TITLE is:

"...4.7 GHz IBM Power6 dual-core processor .."

and the autor of post question about if THIS CPU... it implied a great improvement in chess (he/she asks) ...

This CPU is X2... or X4 more fast that currents CPU... I need do a comparation similar... NOT is perfect OK ¡¡¡ but is BETtER that NOTHING ¡

My opinion is that NOT... THAT it is not the solution... for the time now ...

clear... that if you want to make tables it bases of 32 pieces... uff... pluf.

Ok... NOW... some programers think that chess knowledge in HIS engine are "complete"... ok... test that then whit a FAST CPU... and the other with a slowly CPU... or give more time to analysis...

I estimate... knowledge 70% speed 30%... (if it is not bigger even the difference... ) can that estimates can be made... but makes many many YEARS that ALL people decided that the gross force (speed) it WAS NOT the road for the chess...

Bye , since SPAIN.

Hi Oliver,

If you want to claim that improvements in speed in the last 2 years are less important than improvement in knowledge in the same time then you are probably correct but the claim that speed is less important than knowledge is meaningless when we do not know the difference in speed and the difference in knowledge.

We also do not know the future and we do not know if in the future improvement in speed are going to be less important than improvement in knowledge.

Uri

[Karmazen & Oliver]

Hi Oliver,

If you want to claim that improvements in speed in the last 2 years are less important than improvement in knowledge in the same time then you are probably correct but the claim that speed is less important than knowledge is meaningless when we do not know the difference in speed and the difference in knowledge.

We also do not know the future and we do not know if in the future improvement in speed are going to be less important than improvement in knowledge.

Uri

YEs of course... maybe it´s a good resume.

but the tittle of theme is: "4.7 GHz IBM Power6 dual-core processor"...

I don´t think that THIS CPU will be the "end" of chess ... His speed not it enough for chess... only X2 or X4 or X8 more fast that CPU in this time...

OK. in future the UP of speed maybe helps a lot to "go" more far in the searh tree.. but in this moment the MORE benefits are the algorimts ... that permit CUT a lot of variants not good.

Cristoher say that "a fast" search near end-tables is the most important... to get the "perfect" evaluation. ok.

but in the last moment... the engine must select one variant or other... and when the engine is calculating is HIS selection is good ... spend time... if engine select in fisrt time good variants... need less time to select it...

OK... near end... if a engine find a good line very fast is question or "good" "luck" or other question...

engines arrive very fast in long time controls to PLYS 20-25 plys... this are 12 moves ... maybe 5~7 change pieces on board... or 3~5 ...ok... if end tables are in 7~6 ... when in board "only" are 12 pieces.. some engines have a lot of ventaje if have enough speed to arrive so fast end-tables...

but ? in test actuality ... the end tables only are 3-4-5 pieces... why ? in this test ONLY one engines wins to UP 3000 points ... ELO.

speed ? I don´t think it.

the theme was if THAT CPU IBM 4,7 Ghz are enough SPEED to chess .. I said that NOT because CHESS not is a problem of speed... but helps ¡ obviusly ...

bye since Spain. Oliver