Author: Aaron Gordon
Date: 16:17:07 04/14/03
Go up one level in this thread
On April 14, 2003 at 15:16:13, Tom Kerrigan wrote: >On April 14, 2003 at 07:37:21, Aaron Gordon wrote: > >>On April 14, 2003 at 05:56:53, Jorge Pichard wrote: >> >>>http://www.pcworld.com/news/article/0,aid,110254,00.asp >> >>Of course it is, so was 533 and 400. As they say, supposedly you get 3.2gb/s on >>the P4's "400MHz" bus, 4.2GB/s on the "533MHz" bus and 6.4gb/s on the "800MHz" >>bus. In reality, however, a "533MHz" bus (actually 133MHz) gets about 2.5gb/s. >>4.2Gb/s theoretical, 2.5gb/s actual, thats 59% efficiency. Take the true >>200MHz/400DDR bus of an AthlonXP running on an Nforce2. Whats the Max >>Theoretical you can get? 3.2gb/s... what is the actual bandwidth? 3.2gb/s. Plus >>the memory latency is MUCH lower on an Athlon board. :) > >How are you measuring actual bandwidth? I've never heard of a FSB that has 100% >data utilization. > >-Tom To answer your question real quick, go to http://www.sisoftware.demon.co.uk/sandra/ and grab the latest Sisoft Sandra. Run the memory test and you can compare your actual results with your theoretical bandwidth. Also about the bug in some P4-3.06's.. there is some information on it here: http://www.extremetech.com/article2/0,3973,1022686,00.asp
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.