GitLab

* Top level Python module is now "openmm"

* Updated module names in examples

* Updated module names in documentation

* Updated module in CI scripts

* Added deprecation warning

* Frames distance calculation as matrix multiplciation

* Adds comment explaining distance calculation

* Tunes launch bound for cuda11.2

* Simplifies the effective matrix multiplication
Co-authored-by: David Clark <daclark@nvidia.com>

* Made ordering of impropers more deterministic

* Fixed typo

* Add benchmarks from Amber20 benchmark suite to standard benchmark script

* Add ensemble option; don't change hydrogen mass in amber input files

* Download and extract .tar.gz using pure python code, no wget/tar dependencies

* Rename amber tests

* Initial fixes for Apple silicon

Doesn't work yet

* assume vec4 is supported on apple

* Trust the user with env variables

* Fix macos-version-min logic

* Do not alter mass of hydrogens in rigid water

* Do not alter mass of hydrogens in rigid water

* Update IntegrationUtilities.cpp

* Use pinned buffer and for context download
* use template to reduce duplicate code

* Remove template and only add PinnedBuffer for minor performance improvement

* Use pinned buffer for minor performance improvement

Co-authored-by: xxx <yyy>

* nbthole

* Update charmmpsffile.py

* Update charmmpsffile.py

* Update charmmparameterset.py

fix multiple dihedral with the same periodicity and add a warning reminder for improper dihedral

* add test example for NBTHole

add system files of CYT-GUA-CYT DNA segment for testing NBThole

* Update TestCharmmFiles.py

* Changes name of NVRTC program

* Adds launch bounds for findInteractingBlocks

* Replaces launch bound parameter with named constant
Co-authored-by: David Clark <daclark@nvidia.com>

* Adds cuFFT error checking for execution

* Removes tabs
Co-authored-by: David Clark <daclark@nvidia.com>

* Start branch

* Update installation instructions for conda-forge

* Update instructions for openmm-setup and openmmforcefields

4*(1<<30) evaluates to 0 so the check returns true on 2GB cards. Test runs fine on 2GB cards but requires over 4GB of system memory.

* Added HMR to make ApoA1 benchmark more stable

* Removed debugging code

* Fix errors on Python 3.9

* Revert changes to travis builds