GitLab

Modify Context.getState to accept a set of indices or bitmask

Ensure context can be created when selecting CUDA device

Fixed test failures on OpenCL CPU device

Minor optimizations to CPU neighbor list

Added getNumAtoms, getNumResidues, getNumChains to simtk.openmm.app.Topology

Fix force deepcopy bug

Control what device is used for OpenCL tests

No GMX include directory in example

Fix thread synchronization in customCentroidBond.cl

This change is intended to improve the behavior when multiple
CUDA contexts are created on a node with multiple CUDA GPUs,
each of which is set to process-exclusive mode.

Fix gradient of periodicdistance at 0

Added checks for nan coordinates in CUDA and OpenCL

Add code supporting input with and without units for CharmmPsfFile.