Leveling The Playing Feild

[bob]

I get the impression that a majority of programmers may not like this rule change but a majority of Computer Chess fans will. Time will tell.

I would have thought the majority of programmers would be happy with it actually, most have 8 core machines or less. Computer chess fans seem divided, but as you say time will tell. I'm certainly in favour, I don't like to see 40 cores vs 8 or 40 vs 4. I like to see the talents of the programmers given a chance to shine, with no huge hardware imbalances.

As a matter of interest, did you or any of the Hiarcs team have contact with the ICGA to get this rule introduced ? You've been advocating it for a while I think.

For the record, you have not yet seen a viable 40-core program entered. The 5 x 8 cluster Rybka used was certainly no cause for concern. This is a difficult project to produce reasonable performance on such hardware. The ICGA, which is supposed to "foster computer chess research" is doing exactly the opposite.

Just another nail in the coffin of an already dead event anyway.

[pijl]

I get the impression that a majority of programmers may not like this rule change but a majority of Computer Chess fans will. Time will tell.

I would have thought the majority of programmers would be happy with it actually, most have 8 core machines or less. Computer chess fans seem divided, but as you say time will tell. I'm certainly in favour, I don't like to see 40 cores vs 8 or 40 vs 4. I like to see the talents of the programmers given a chance to shine, with no huge hardware imbalances.

It does require talents to be able to use all that processing power to the benefit of a chess program. It is no challenge to get many nps, it is a challenge to use them effectively.
I'm one of the programmers that do _not_ like this rule. And I will probably not use more than 4 cores in the next tournaments either. It's a matter of principle. If you succeed in taming all those CPUs, you deserve the credit for it. It is not an easy task.

But what is a core, exactly? I believe that most 'cores' are also capable in handling multiple instructions at the same time (4 in the popular Intel's I believe). Do hyperthreading capabilities count? What about GPU's? Windows does not report them as CPUs, so you can use as many of those as you like? Suppose someone tries the FPGA method again or something similar. Is that a core as well?

And what measure is next ... Prescribing the opening positions from which you have to play to diminish the effect that professional book authors have on the tournament?

I am not a favorite to win the WCCC. This rule does not change that, nor does it change it for most others. The influence of hardware on the winner of the WCCC is exaggerated.

Richard.

[Dann Corbit]

I don't know about you, but I feel it's about time they started to level the playing field hardware wise. Things were getting out of hand. I would not call 4 cores vs 20 cores a fair match.

Quote From: David Levy

"The ICGA feels that the time has come to take stock of this trend of hunting for astronomic numbers of cores, and to bring matters back to Earth. We see no point in organizing an event that can be won by a simple weight of processing power, when just about all the competing programs are able to use computers with only a handful of processors. By allowing 20 cores, or 40, or 80, at the present time, we would be saying to the vast majority of chess programmers that, if they want their program to be able to give of its best, they must first acquire the use of an expensive computer system with a very large number of processors. That is not what we believe the World Computer Chess Championship should be about. One should not be able to buy the title in this way."

Read More:
http://www.hiarcs.net/forums/viewtopic.php?t=2008

Marvelous. Rules that have been used since the first ACM CC event in 1970, and the first WCCC in 1974 are now "not good enough"???

If they want to have a uniform platform event, that's fine. But why start to limit the WCCC hardware? They already have difficulties getting 16 programs to show up. Are they trying to reduce the field to 4 or something?

I think that uniform platform is largely unachievable. What if a program is designed for Apple Mac? What if Thinker wants to enter on a mobile phone (which will still be a strong opponent)?

And of course, it would exclude Hydra or other super-strong entrants from applying to play.

If you could get the entire Internet to supply its compute power to play chess, I would like to see it.

But the ICCA can do whatever they want. I think by now most people have stopped caring what they do.

[Nimzovik]

Indeed an unfair advantage there may be with an abundance of processors in relation to the opposition. However the qulity of games would go up --yes? Truth on the board is what I like to look for.....

[CRoberson]

I disagree with the idea all together.

There are teams that enter WCCC that regularly use their contacts
inside of Intel to get hardware that you can't buy. That will certainly
give them an edge with such a new rule.

On the topic of uniform platform events: what do you call SSDF, WBEC,
RWBC, CCRL and so forth.

The reason the public can buy multi-core/thread/process software
is because there are events where people can push the limits. Without
such events, the progress would be impeded.

[Dirt]

And what measure is next ... Prescribing the opening positions from which you have to play to diminish the effect that professional book authors have on the tournament?

That's one possibility. Another would be to require all the teams to use the same chess engine so that no book maker would have an advantage over the others.

[Dann Corbit]

And what measure is next ... Prescribing the opening positions from which you have to play to diminish the effect that professional book authors have on the tournament?

That's one possibility. Another would be to require all the teams to use the same chess engine so that no book maker would have an advantage over the others.

If you use the same engine and alter the opening books, you will get a tournament like this one:
http://www.sedatchess.com/scct4.html

[Zach Wegner]

I am very much against this. There is currently only one motivation pushing forth research into utilizing large degrees of computer power in chess, and that is a few tournaments a year that allow it in competition. And research in this area is useful and important in my opinion. There are plenty of other avenues for encouraging the development of 32 bit single processor ponder off learning off vanilla chess strength. Those few competitions that encourage anything goes pursuit of chess perfection are rare and precious.

-Sam

Indeed, and I will add that not only does it take away incentive, but it takes away opportunity. Those that were able to play on supercomputers in the past (Zappa and Diep being the most recent examples) were given the opportunity because of the WCCC. SGI or some similar company isn't going to give up valuable supercomputer time for something like CCT, especially when the WCCC exists. So those that do want to research it are left with virtually no options, excepting Bob with his personal cluster. This rule is completely preposterous and anticompetitive, and I dare say was created with the only intent of handicapping Rybka.

[lexdom]

A tournament with unlimited hardware and another with limited hardware similar to WCCC and WMCCC?

[bob]

I am very much against this. There is currently only one motivation pushing forth research into utilizing large degrees of computer power in chess, and that is a few tournaments a year that allow it in competition. And research in this area is useful and important in my opinion. There are plenty of other avenues for encouraging the development of 32 bit single processor ponder off learning off vanilla chess strength. Those few competitions that encourage anything goes pursuit of chess perfection are rare and precious.

-Sam

Indeed, and I will add that not only does it take away incentive, but it takes away opportunity. Those that were able to play on supercomputers in the past (Zappa and Diep being the most recent examples) were given the opportunity because of the WCCC. SGI or some similar company isn't going to give up valuable supercomputer time for something like CCT, especially when the WCCC exists. So those that do want to research it are left with virtually no options, excepting Bob with his personal cluster. This rule is completely preposterous and anticompetitive, and I dare say was created with the only intent of handicapping Rybka.

What is funny is that they were handicapping a program that didn't need handicapping. First, Rybka is the strongest thing around anyway. Second, the "cluster rybka" is no stronger than the normal single-machine rybka, but it sounds better.

I've already decided that I would not participate in another WCCC event anyway, so I really don't care. If they want to hasten their own demise, more power to 'em...