alphataubio

@akohlmey has anybody done anything on this issue yet ? id be willing to pitch in next if you give me some initial guidance on what path to follow. we...

Ok I'll reverse commit and close PR 4035

My experience with GPUs has been similar in many cases... you gain an order of magnitude by vectorizing a math operation in the GPU but you lose two orders of...

98% tests passed, 13 tests failed out of 573 The following tests FAILED: 452 - FixTimestep:deform (Failed) 454 - FixTimestep:deform_tri (Failed) 456 - FixTimestep:efield_const (Failed) 457 - FixTimestep:efield_region (Failed) 460...

there's no rush to merge this (NOT READY). i improved the unit test coverage for KOKKOS from 34% to 43% but there are now tests failing so it should NOT...

> Nice! until you find out that 13 tests are now failing ...

Also i ran the unit tests with GoogleTest [Undefined Behavior Sanitizer](https://clang.llvm.org/docs/UndefinedBehaviorSanitizer.html), [Address Sanitizer](https://github.com/google/sanitizers/wiki/AddressSanitizer), and [Thread Sanitizer](https://github.com/google/sanitizers/wiki/ThreadSanitizerCppManual). **No issues detected** see https://github.com/lammps/lammps/issues/4229#issuecomment-2268817624

@stanmoore1 i tested `fix efield/kk` on both 1 and 4 GPUs. It works, I was concerned about comment in the code about crash accessing unwrap[] but it doesn't and numerics...

> @alphataubio overall this looks good. I had one question in the comments. No I don't need fsum in Kokkos subclass anymore Cleanup commits done, unit tests all pass

It still needs a review by @sjplimp because I added functions to base classes but no rush next stable is not for another year. If @stanmoore1 you see people having...