Search found 2821 matches

by Rebel
Fri Apr 06, 2012 4:53 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash table division
Replies: 42
Views: 6012

Re: Hash table division

Houdini wrote:
Rebel wrote:Of course I agree with you that a faster search should be the end result and for me it does, I am just trying to explain the logic behind the approach.
Why does it result in a faster search?
Because of testing, what else is the final measurement?
by Rebel
Fri Apr 06, 2012 2:22 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash table division
Replies: 42
Views: 6012

Re: Hash table division

When implementing hash tables in the days that memory was limited (1 Mb if you were rich) I set-up my hash table data structure as follows: - One hash table for odd plies - One hash table for even plies And where ever possible make the odd ply HT twice as big as the even ply HT because that one is ...
by Rebel
Fri Apr 06, 2012 2:10 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash table division
Replies: 42
Views: 6012

Re: Hash table division

AFAIK i was the first one really using 4 sequential or more probes. Bob described in 90s doing 8 random probes (to avoid chaining), but of course at modern memory systems using more than 1 table is pretty slow and/or gives some overhead. Yes, I think we all have expirimented with the optimal number...
by Rebel
Fri Apr 06, 2012 2:06 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash table division
Replies: 42
Views: 6012

Re: Hash table division

And so it makes sense to declare a bigger odd-ply-HT than an even-ply-HT. Not really. Number of stores in meaningless. Number of hits is only a little bit less meaningless. What really counts is how much search time the hits save you. When you say 'odd ply', do you mean counted from the root or cou...
by Rebel
Fri Apr 06, 2012 1:53 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash table division
Replies: 42
Views: 6012

Re: Hash table division

AFAIK i was the first one really using 4 sequential or more probes. Bob described in 90s doing 8 random probes (to avoid chaining), but of course at modern memory systems using more than 1 table is pretty slow and/or gives some overhead. Yes, I think we all have expirimented with the optimal number...
by Rebel
Fri Apr 06, 2012 1:29 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash table division
Replies: 42
Views: 6012

Re: Hash table division

When implementing hash tables in the days that memory was limited (1 Mb if you were rich) I set-up my hash table data structure as follows: - One hash table for odd plies - One hash table for even plies And where ever possible make the odd ply HT twice as big as the even ply HT because that one is ...
by Rebel
Thu Apr 05, 2012 8:30 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash table division
Replies: 42
Views: 6012

Hash table division

When implementing hash tables in the days that memory was limited (1 Mb if you were rich) I set-up my hash table data structure as follows: - One hash table for odd plies - One hash table for even plies And where ever possible make the odd ply HT twice as big as the even ply HT because that one is a...
by Rebel
Wed Apr 04, 2012 6:18 pm
Forum: Computer Chess Club: General Topics
Topic: Question for Ed Schröder about Nona
Replies: 3
Views: 1278

Re: Question for Ed Schröder about Nona

Hi Gerd, Frans ported his engine (Nona) to my interface code for the Mephisto hardware because my hardware was running at a higher speed and at the time his program was considered stronger. The idea was also to commercialize it but as far as I can remember that never happened. Hi Ed, I see, interes...
by Rebel
Tue Apr 03, 2012 8:53 pm
Forum: Computer Chess Club: General Topics
Topic: Question for Ed Schröder about Nona
Replies: 3
Views: 1278

Re: Question for Ed Schröder about Nona

Hi Gerd,

Frans ported his engine (Nona) to my interface code for the Mephisto hardware because my hardware was running at a higher speed and at the time his program was considered stronger. The idea was also to commercialize it but as far as I can remember that never happened.
by Rebel
Tue Apr 03, 2012 1:22 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Elo versus speed
Replies: 29
Views: 7956

Re: Elo versus speed

It would also be interesting to compare ratings of parallel speedups for 1,2 and 4 processors. Incidentally this seems to roughly match your tests if you assume a 1,1.8 and 3 speedups. That is generally expected efficiency for YBW implementation. The increase in elo is much lower as expected. For e...