Sergey Klevtsov

I think this is as ready as it's going to be. Of the three failing tests that remain, two aren't expected to pass and one only has tiny diffs (see...

We'll need to implement a spack package recipe for fesapi, but it can be done after the initial integration into GEOSX is complete IMO. For now, to avoid breaking spack...

> To avoid too many conflicts until this draft will be ready, I suggest to turn this pull request into three ones: > > * 1st one: extract vtk logic...

This is specific to Acoustic solver due to some hard coding going on here: https://github.com/GEOSX/GEOSX/blob/d3d351883195c08686ffff438957fe59a6d916e8/src/coreComponents/physicsSolvers/wavePropagation/AcousticWaveEquationSEM.cpp#L471

> The error can be seen here: https://github.com/GEOS-DEV/GEOS/actions/runs/7674651265/job/20919616872?pr=2873 I don't see any erros in that log that have anything to do with `array`? There's a bunch of deallocation errors for...

Note: code in GEOSX that uses this is currently disabled behind an `#if 0`, because I found out some Tpetra-based solvers only work reliably with Unified Memory, and so the...

> Do we need to update Trilinos? No, there's no new version available. And it's not something they seem to have fixed yet. All Tpetra build instructions, [FAQs](https://github.com/trilinos/Trilinos/blob/688a92473236fc2103a77690cda7265a47dc6f5f/packages/tpetra/doc/FAQ.txt#L94-L106), etc. explicitly...

Hi @liushilongpku, If you don't mind, can you post the output of `nvidia-smi`? Also, can you try the following in you WSL console (assuming you have `python` and `pip` installed):...

> cuda-python(but not python-cuda) Apologies, shouldn't have typed it from memory :) Corrected my previous message for history. It seems _hypre_ is allocating its device memory as unified, the error...