[petsc-dev] Mat assembly on GPUs ...

[Mark Adams]

On Sun, Mar 29, 2020 at 7:51 PM Jed Brown <jed at jedbrown.org> wrote:

> Mark Adams <mfadams at lbl.gov> writes:
>
> > On Sun, Mar 29, 2020 at 6:20 PM Jed Brown <jed at jedbrown.org> wrote:
> >
> >> We are interested in assembling point-block diagonals for use in
> >> pbjacobi smoothing.  This is much simpler than AIJ,
> >
> >
> > Would the coloring/DD algo on the outside be the same, just with a
> > different assembly kernel?
>
> In one mode, though we can also do analytic stuff on the inside to avoid
> coloring (in which case we produce a bunch of elements or block-diagonal
> blocks).
>
> > Or are you looking a fusing the element construction with assembly? That
> > seems reasonable now that I think about it.
> >
> > It's the same coloring, right?
> >
> > And, do we have an appropriate coloring method now?
>
> Are you asking for a GPU-native coloring?


No. And I am using p4est grids.


> Does your nonzero pattern
> change frequently?
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20200329/729977b4/attachment.html>