Author: Uri Blass
Date: 08:12:38 11/02/05
Go up one level in this thread
On November 02, 2005 at 10:57:35, Gregory Owett wrote: >Recently, I launched two tournaments (ponder off) with the same participants, on >two different machines. Here results: > >On P4 2.2 T=15'+5" : > >1. Fruit 2.2.1 KS104 7,5 / 10 >2. Shredder 9 UCI 5,5 >3. Shredder 7.04 CLm1 5 25,00 >4. Fruit 2.2.1 Uri 5 23,00 >5. Shredder 8 CLD 3,5 17,25 >6. Fruit 2.2.1 3,5 16,50 > > >On AMD x2 dual core 4800+ T=15'+5" : > >1. Shredder 8 CLD 6,5 / 10 >2. Fruit 2.2.1 Uri 6 >3. Fruit 2.2.1 KS104 5 25,75 >4. Fruit 2.2.1 5 24,25 >5. Shredder 9 UCI 4,5 >6. Shredder 7.04 CLm1 3 > > >The success of Shredder 8 CLD (UCI) on the dual core machine, confirms what the >task manager indicated, i.e. which Shredder 8 CLD, in spite of it is >single-threaded engine, used 99% of the two processors, while all the others >used 50%, in other words, only one processor. results may be because of luck so they do not prove nothing. The way to check it is to look at the number of nps. I am also not sure if Shredder is a single threaded engine. Shredder9 is only for single processor but I remember that at some point of time there was no deep shredder and shredder was SMP capable. > >(Native engines, as F8, J9, H9, CT15 etc...use 99% too.) No by definition they cannot do it because they were not programmed to do it. Hiarcs and tiger always supported only a single processor so claiming that they use more than 50% cannot be correct. 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.