Author: Uri Blass
Date: 09:40:15 01/11/01
Go up one level in this thread
On January 10, 2001 at 23:37:45, Aloisio Ponti Lopes wrote: >On January 10, 2001 at 10:25:30, Hermano Ecuadoriano wrote: > >>If it is weaker than F6 on one processor, the following ad >>http://www.chessbase.com/newproducts.htm#deep >>certainly doesn't warn the customer of that fact. > >Ok, I can read this at the above link: > >"DEEP FRITZ Multiprocessor Version >Deep Fritz is the multiprocessor version of Fritz and supports one to eight >processors. On a dual processsor machine, it will run about 80% faster than on a >single." > >That is a very clear statement. ChessBase says it supports a single processor, >but that it was made as a multiprocessor version (the purpose of the program is >to run on SMP machines). Being 80 % faster on a dual than on a single, why >should it be tested on a single processor? Just to see that it really runs 80% >slower? Isn't faster better when we talk about programs? (ok, that's another >poll- is faster better or not?) Chessbase wanted Deep fritz to be tested on a single machine. I do not see what is your problem. Why do you think that the ssdf should be against chessbase and refuse to test Deep Fritz on a single processor. >This testing is nonsense, IMHO. >That's the same thing of testing a Winboard engine using Shredder's GUI- as we >know it has a bug, so the test must wait for the patch. It is not the same. We know that deep Fritz is better with more processors but the question is about Deep Fritz's rating on a single processor. Testing winboard engines under Shredder is different because the people who have the winboard engines can use them not under shredder without paying maney for it. SSDF should test dual >versions on dual machines. That's my point of view. Clear and simple. Test dual >versions on dual machines. If SSDF don't have dual machines, that's not my >fault. Most of us don't have a dual. What about testing it against those >single-processor programs using a dual or a quad or a 8-processor machine? >Wouldn't that hurt the single-processor program (that can't take advantage of >more processors)? So the statement could be different... "WOW! Deep (anything) >just beat single (something) , e.g.: 20 - 4 !!!! on a dual!". > >Pure nonsense, IMHO. The testing is not nonsense. It is relevant for customers who have only one procesor. I bought Deep Fritz because I believed that it is better than Fritz6a also on a single processor. Chessbase claimed that it is the case and I believed them because I read good results of Deep fritz on one processor before the ssdf games and I also assumed that chessbase decided to let the ssdf to test Deep Fritz because they know that it is better than Fritz6a. I have not machine with more than one processor and I did not decide if I am going to buy a machine with more than one processor and when I am going to do it. I think that the ssdf results may be relevant for other people in order to decide if to buy Fritz6 or Deep Fritz. Uri
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.