Re: Performance--Dual Core
- To: mathgroup at smc.vnet.net
- Subject: [mg70123] Re: [mg70089] Performance--Dual Core
- From: "Christoph Lhotka" <lhotka at astro.univie.ac.at>
- Date: Wed, 4 Oct 2006 05:59:39 -0400 (EDT)
Hi ! When using Windows you have to tell the OS the following: > ---------------------------------------- > Variable Name : OMP_NUM_THREADS > Variable Value: % NUMBER_OF_PROCESSORS% (e.g. 2) > ---------------------------------------- (first create a new variable OMP_NUM_THREADS as a System Variable in the Device Manager, then set the NUMBER_OF_PROCESSORS to 2 (or 3, ...). The values of OMP_NUM_THREADS should be equal to the value of NUMBER_OF_PROCESSORS... After doing so, you will see, that Mathematica will use both cores for its computation... Christoph On Tue, 3 Oct 2006 06:16:29 -0400 (EDT) oparadis at princeton.edu wrote: > I have just bought a new system (AMD 64 X2 Dual Core 4200+) thinking it > would improve computations with mathematica 5.2. I am mainly running > symbolic computation problems (Solve,Reduce,FullSimplify stuff) involving a > system of nonlinear equations with domain restrictions over parameters... > > I am puzzled by the fact mathematica seems to be using the CPU resources > from a single core at the time even though the problem it tries to solve is > very complicated (takes 1 day approx.). While computing, windows' task > manager shows "MathKernel" using 50% of CPU and "System Idle Process" using > the remaining 50%. > > I have 2 questions. First, what's going on ? Is Mathematica really unable > to use both cores at the same time ? Second, how can one deal with this ? > Is there a way to force mathematica to somehow split the task between both > cores ? > > I know some of the previous posts have dealt with similar issues but they > don't quite seem to answer my specific problem. > > > > Link to the forum page for this post: > http://www.mathematica-users.org/webMathematica/wiki/wiki.jsp?pageName=Special:Forum_ViewTopic&pid=14120#p14120 > Posted through http://www.mathematica-users.org [[postId=14120]] > > -- Mag. Christoph Lhotka -- University of Vienna / Department of Astronomy Tuerkenschanzstrasse 17, A-1180 Vienna, Austria fon. +43 (1) 4277 51841 mail. lhotka at astro.univie.ac.at