Author: Vincent Diepeveen
Date: 07:45:12 10/31/02
Go up one level in this thread
On October 30, 2002 at 17:56:50, Dieter Buerssner wrote: then not fulltime, but your thing was so much faster on 16 bits cpu's than on 32 bits cpu's that i was really amazed ;) Then i learned you owned a K6 and the thing was clear to me :) But in case of frans it is well known he always says that each new version of him he completely rewrites for a new cpu. That is a fulltime job. What i meant to say is that you care shit usually for 8 bits versus 32 bits mixing as long as it is faster for you you go for it. If that works for the cpu you own now (k7 hopefully now) it might work bad on a new K7 design or at a new P4 core ;) >On October 29, 2002 at 19:55:08, Vincent Diepeveen wrote: > >>Optimizing for a certain processor like Dieter and Frans and many >>others (Ed!) always use(d) to do is really a fulltime job IMHO. > >If you mean me, by "Dieter": I beg your pardon. It is not true that I am >optimizing for one specific processor. On the contrary, I try to only use >portable ISO C. My engine worked well on Atari ST, VAX, some MIPS (under IRIX >and VMS), x86, and probably more, I forgot at the moment. Before I included the >TB-access code of Eugene Nalimov, it even worked well with 16 bit compilers. My >engine at the moment does not use any inline assembler. Perhaps some high level >constructs are inspired by looking at assembly output of some compilers - but >not only from one platform. > >I never did a "fulltime job" for my chess engine ... > >Regards, >Dieter
This page took 0 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.