Matthew Martineau
Matthew Martineau
Just to add some additional insight here. The determinism_flag applies to aggregators and matrix coloring (so specific to the aggregation based AMG rather than classical). Still, I think we should...
I'm hopefully fixing this in #184
Yes that's exactly correct - the current version is just hacked but it shows the desired outcome of having only compile time type selection. CMake flags like you suggested will...
If you mean generally for AmgX then, yes, I definitely do - we have been considering it for a while.
Both the setup and solve phases in AmgX are performed on the GPU unless you tell AmgX to run on the host. It is straightforward to run with GPU resident...
The distributed model for this application leverages MPI and that is currently the standard option for such problems. You can domain decompose your problem, pass it to AmgX and the...
Yes it's a nice pattern from the perspective of performance - you should be able to just upload a new b with AMGX_vector_upload and call AMGX_solver_solve_with_0_guess again. Essentially, if A...
Looks good - thanks!
I took a look to see if I could find why this was designed this way in the first place, and it seems to be a purposeful API design decision....
If we are happy to store a temporary copy of the handle when MPI is enabled then this would be an ideal solution. In my mind though changing whether we...