Author: Vincent Diepeveen
Date: 10:22:30 06/05/02
Go up one level in this thread
On June 05, 2002 at 04:14:41, Gerd Isenberg wrote:
Gerd, a 33% speedup isn't much if you first slow down 2 times.
Right now the crafty datastructure is exactly 2 times slower doing
what i can do without bitboards.
Hammer needs its own compiler too. We will see it come in time.
Let's hope it can do cheap more than dual also.
It is time there come processors which can get parallellized with
shared memory very easily. Even clustering them with very fast
access times would be interesting. Right now such switches and network
cards that do 0.5 usec latency (or less) are simply too expensive.
The only advantage from bitboarders is they have no clue what making
an evaluation is.
>IMHO a 33% speedup over nonbitboards is a lot, because also bitboard progams do
>also a considerable amount of time of their search, movegen and eval with none
>bitboard stuff. But we'll see, hammer is not so far away.
>
>In functions like getting all pieces controlling a square, i expect a speedup of
>more than 100% with hammer.
>
>BitBoard CNode::GetAttackedBy(unsigned int sq) const
>{
> return
> (RookAttacks(sq) & RookMover())
> | (BishopAttacks(sq) & BishopMover())
> | (BPawnAttacks(sq) & GetWPawnBB())
> | (WPawnAttacks(sq) & GetBPawnBB())
> | (KnightAttacks(sq) & GetKnightBB())
> | (KingAttacks(sq) & GetKingBB())
> ;
>}
>
>
>First the code length is about the half or less, because of only one mov-,
>and-, or-instruction per bitboard instead of two.
>
>Second, hammer has more general purpose registers than the anachronistic x86, so
>there is no need here to store temporary resulst in memory.
>
>The potential to use multiple integer pipes is nevertheless quite good here.
>
>Gerd
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.