Mateus R 06d8d51329 Amortize PME atom-grid re-sort for large systems (#5305) ...

Large systems (>15000 atoms) re-sorted the PME atom grid every step,
skipping the step-counter amortization used for smaller systems. On
current GPUs the per-step sort is mostly wasted work, so re-sort every
2 steps instead. Smaller systems are unchanged. The sort only changes
charge-spread memory locality; results are identical up to floating-point
summation order.

2026-05-29 17:52:44 -07:00

..

common

Amortize PME atom-grid re-sort for large systems (#5305)

2026-05-29 17:52:44 -07:00

cpu

atomFlags switched to unordered_map for performance (#5247)

2026-03-31 16:46:07 -07:00

cuda

Make sure contexts are deselected before evaluation (#5279)

2026-05-04 16:26:18 -07:00

hip

Handle hipErrorFileNotFound gracefully in ROCm 7 (#5303)

2026-05-29 10:43:07 -07:00

opencl

Compile correctly under newer C++ standards (#5296)

2026-05-27 11:01:05 -07:00

reference

More barostats support scaling particles independently (#5278)

2026-05-26 07:37:26 -07:00