Please be more precise with your comments. Git works just fine on KNL and
I have hundreds of data to confirm this. Any issues you have with Git on a
KNL-based platform are the result of the software configuration.
Furthermore, every Haswell Xeon binary runs on KNL* so there is no
hardware-related
Matthew Knepley writes:
> I took a look at the code. It is the worst kind of C-in-C++, It would
> probably take 1 day to make a pure C version :)
> The claim from Edgar is that it is intended to be used for this purpose
> (have you read the arXiv thing with Torsten
> about sparse tensor computing?
On Sat, Feb 18, 2017 at 3:38 PM, Jed Brown wrote:
> Matthew Knepley writes:
>
> > Have we considered using this
> >
> > https://github.com/solomonik/ctf
> >
> > as an alternative sparse matrix-matrix code for GAMG? Or was it
> unsuitable
> > for some reason.
>
> I don't think we considered it
Matthew Knepley writes:
> Have we considered using this
>
> https://github.com/solomonik/ctf
>
> as an alternative sparse matrix-matrix code for GAMG? Or was it unsuitable
> for some reason.
I don't think we considered it for that purpose. I'm not sure if its
dynamicism would work against us
Have we considered using this
https://github.com/solomonik/ctf
as an alternative sparse matrix-matrix code for GAMG? Or was it unsuitable
for some reason.
Thanks,
Matt
--
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than