[petsc-users] MatMatSolve for MUMPS disappeared?

[Jed Brown]

On Wed, Sep 19, 2012 at 6:43 AM, Alexander Grayver
<agrayver at gfz-potsdam.de>wrote:

> On the order of 10^6 so far.
>

So 10^3 vectors each of size 10^6 is already several GB which is larger
than local (NUMA) memory for a cluster node (and larger than the entire
memory of some nodes). The spill on rank 0 will cause lots memory allocated
by rank 0 later to spill into the memory bus/NUMA region of other
sockets/dies of the shared memory node (think of a four socket system, for
example). This can easily slow the rest of your program down by a factor of
3 or more.


> I guess memory has never been a problem because internally PETSc used
> MatMatSolve_Basic for MatMatSolve with MUMPS. Thus you never gather more
> than one rhs.
>

Yup, likely.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20120919/cacb1f6f/attachment.html>