[petsc-dev] PETSc programming model for multi-core systems

[Rodrigo R. Paz]

Hi, Aron,
IMHO, the main problem here is the low memory bandwidth (FSB) in Xeon E5420
nodes.
In sparse matrix-vector product, the ratio between floating point operations
and memory accesses is low. Thus, the overall performance in this stage is
mainly controlled by memory bandwidth.
We have also tested on i7 node (QPI  memory controller) and results are
not appreciably improved (see the right plot of the attached figure).

Rodrigo

--
Rodrigo Paz
National Council for Scientific Research CONICET
CIMEC-INTEC-CONICET-UNL.
Güemes 3450. 3000, Santa Fe, Argentina.
Tel/Fax: +54-342-4511594, Fax: +54-342-4511169


On Fri, Nov 12, 2010 at 10:32 AM, Aron Ahmadia <aron.ahmadia at kaust.edu.sa>wrote:

> Hi Rodrigo,
>
> These are interesting results.  It looks like you were bound by a
> speedup of about 2, which suggests you might have been seeing cache
> capacity/conflict problems.  Did you do any further analysis on why
> you weren't able to get better performance?
>
> A
>
> On Fri, Nov 12, 2010 at 8:26 AM, Rodrigo R. Paz
> <rodrigop at intec.unl.edu.ar> wrote:
> > Hi all,
> > find attached a plot with some results (speedup) that we have obtained
> some
> > time ago with some hacks we introduced to petsc in order to be used on
> > hybrid archs using openmp.
> > The tests were done in a set of 6 Xeon nodes with 8 cores each. Results
> are
> > for the MatMult op in KSP in the context of the solution of
> > advection-diffusion-reaction eqs by means of SUPG stabilized FEM.
> >
> > Rodrigo
> >
> > --
> > Rodrigo Paz
> > National Council for Scientific Research CONICET
> > CIMEC-INTEC-CONICET-UNL.
> > Güemes 3450. 3000, Santa Fe, Argentina.
> > Tel/Fax: +54-342-4511594, Fax: +54-342-4511169
> >
> >
> > On Thu, Nov 11, 2010 at 10:34 PM, Barry Smith <bsmith at mcs.anl.gov>
> wrote:
> >>
> >> On Nov 11, 2010, at 7:22 PM, Jed Brown wrote:
> >>
> >> > On Fri, Nov 12, 2010 at 02:18, Barry Smith <bsmith at mcs.anl.gov>
> wrote:
> >> > How do you get adaptive load balancing (across the cores inside a
> >> > process) if you have OpenMP compiler decide the
> partitioning/parallelism?
> >> > This was Bill's point in why not to use OpenMP. For example if you
> give each
> >> > core the same amount of work up front they will end not ending at the
> same
> >> > time so you have wasted cycles.
> >> >
> >> > Hmm, I think this issue is largely subordinate to the memory locality
> >> > (for the sort of work we usually care about), but the OpenMP could be
> more
> >> > dynamic about distributing work.  I.e. this could be an OpenMP
> >> > implementation or tuning issue, but I don't see it as a fundamental
> >> > disadvantage of that programming model.  I could be wrong.
> >>
> >>   You are probably right, your previous explanation was better.  Here is
> >> something related that Bill and I discussed, static load balance has
> lower
> >> overhead while dynamic has more overhead. Static load balancing however
> will
> >> end up with some in-balance. Thus one could do an upfront static load
> >> balancing of most of the data then when the first cores run out of their
> >> static work they do the rest of the work with the dynamic balancing.
> >>
> >>   Barry
> >>
> >> >
> >> > Jed
> >>
> >
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20101112/d0877e00/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: speedup-matsol-nikola.pdf
Type: application/pdf
Size: 48063 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20101112/d0877e00/attachment.pdf>