Memory
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Raghubhushan Pasupathy rpasupathy at hotmail.comMon Jul 17 10:16:50 PDT 2000
- Previous message: philosophical question - limits of beowulf defn
- Next message: philosophical question - limits of beowulf defn
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Hello Folks, I am running a sixteen processor (8 duals with SMP, 256MB memory) cluster in the School of Civil Engineering, Purdue. Here is my problem: I run a Fortran program sequentially on a single processor and I am able to define more than 30 arrays, 900,000 elements (integer) in size without the machine swapping. In fact, this is still not my upper limit. However, when I run this code parallely (MPI) on two processors in the same machine, I am unable to go beyond a 6000 array size. I get a "signal 11" and much before that the machine starts swapping. My question is: The numbers do not add up at all. Even if I duplicate all the arrays for the second process, the machine should easily be able to handle more than 30 arrays of 6000 elements. I do not understand what is going on. Any help would be much appreciated. Raghu Raghu Pasupathy School of Civil Engineering Purdue University West Lafayette, IN 47906 (765) 494 2206 (off) (765) 495 6470 (res) rpasupathy at hotmail.com ________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com
- Previous message: philosophical question - limits of beowulf defn
- Next message: philosophical question - limits of beowulf defn
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
