Author: Robert Hyatt
Date: 21:23:52 12/10/02
Go up one level in this thread
On December 10, 2002 at 23:05:22, Jeremiah Penery wrote: >On December 10, 2002 at 22:51:59, Robert Hyatt wrote: > >>Several post one set of results. One person posts another. And _that_ >>post is the "bible"??? > >No, but neither is the post totally worthless as you claim it to be. I've come to the conclusion that _most_ of his posts are worthless. A few examples: 1. The xeons with hyper-threading are not available and probably won't be available for a year or two. Even though I just bought one from Dell and have it in my office. 2. hyper-threading doesn't work. Even though both I and that "ferkin idiot" nalimov both posted results showing it _does_ work. 3. parallel search is deterministic in its behavior. Even though every parallel researcher on the planet has discussed this non-deterministic behavior and how annoying it can be. 4. speedups > 2.0 for two processors and > 4.0 for four processors can be done and he "proofed" it. Even though any textbook discussing parallel algorithms totally debunks this myth right off the bat... 5. If a program isn't "diep" it is "crap" or "buggy" or "rude" or "simple" or whatever else you want... even though he can't reliably beat those programs with his. 6. The Intel compiler is buggy and can't produce correct code, even though it is used by professionals all over the world. 7. Crafty produces _no_ speedup when running on a dual processor. Even though everyone that tried it came up with numbers close to the 1.7X I have mentioned for several years. I will get tired of typing before I run out of examples... The list goes on and on... crap, crap and more crap. Covered with crap. With crap on the top.
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.