<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;">Thank You for Your answers.<br><br>I have another question<br><br>I wanted to try a 64 bit system out and got the ubuntu with 2 GB RAM and additional 11 GB swap and still I get memory allocation error like this:<br><br><span style="font-style: italic;">[0] Maximum memory PetscMalloc()ed 2,514,564,304 OS cannot compute size of entire process</span><br style="font-style: italic;"><br>Do I miss something when I compile the binaries?<br><br>Regards<br><br>--- <b>18.5.11 (Śr), Barry Smith <i><bsmith@mcs.anl.gov></i></b> napisał(a):<br><blockquote style="border-left: 2px solid rgb(16, 16, 255); margin-left: 5px; padding-left: 5px;"><br>Od: Barry Smith <bsmith@mcs.anl.gov><br>Temat: Re: [petsc-users] Matrix stored on many machines, filled by one process<br>Do: "PETSc users list" <petsc-users@mcs.anl.gov><br>Data: 18 Maj 2011 (Środa),
23:32<br><br><div class="plainMail"><br>On May 18, 2011, at 4:05 PM, Matthew Knepley wrote:<br><br>> On Wed, May 18, 2011 at 3:54 PM, Bartłomiej W <<a ymailto="mailto:bartlomiej.wach@yahoo.pl" href="/mc/compose?to=bartlomiej.wach@yahoo.pl">bartlomiej.wach@yahoo.pl</a>> wrote:<br>> Hello again,<br>> <br>> I am wondering if it is possibile to have a matrix be stored across several machines but have only one process to set its values? Can I just create the matrix and then simply setownership range for process 0 to full size of the matrix and zero for others?<br>> <br>> You can create a parallel matrix, and then only set values from 1 process. Its not efficient, but it will work.<br><br> For big problems it will be terribly slow, so slow as to be idiotic.<br><br> Note in our users manual introduction we specifically say <br><br>PETSc should not be used to attempt to provide a “parallel linear
solver” in an otherwise sequential<br>code. Certainly all parts of a previously sequential code need not be parallelized but the matrix<br>generation portion must be to expect any kind of reasonable performance. Do not expect to generate<br>your matrix sequentially and then “use PETSc” to solve the linear system in parallel.<br><br><br>> <br>> Matt<br>> <br>> Thank You for an answer.<br>> <br>> <br>> <br>> -- <br>> What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>> -- Norbert Wiener<br><br></div></blockquote></td></tr></table>