Commits · 4949017bd2594295f12f8b34d72625dbb09dec9c · tsoc / openmm

27 Jul, 2016 2 commits
- Continuing CUDA implementation of parameter derivatives · 4949017b
  Peter Eastman authored Jul 27, 2016
  
  4949017b
- Began CUDA implementation of parameter derivatives · eae8def5
  Peter Eastman authored Jul 27, 2016
  
  eae8def5
06 Jul, 2016 2 commits
- Continuing to implement derivatives with respect to parameters · 8e01339d
  peastman authored Jul 06, 2016
  
  8e01339d
- Beginnings of support for derivatives with respect to parameters · cd566c63
  peastman authored Jul 06, 2016
  
  cd566c63
25 May, 2016 1 commit
- Improved method for resizing neighbor list · 6cc2edd1
  Peter Eastman authored May 25, 2016
  
  6cc2edd1
05 May, 2016 1 commit
- Changing box vectors does not change positions · 32ea2e9d
  Peter Eastman authored May 05, 2016
  
  32ea2e9d
21 Apr, 2016 1 commit
- Unified names of platform properties · f12cb008
  Peter Eastman authored Apr 21, 2016
  
  f12cb008
29 Mar, 2016 1 commit
- Added property to request deterministic forces · cddeb60f
  Peter Eastman authored Mar 28, 2016
  
  cddeb60f
28 Mar, 2016 1 commit
- Fixed bug using tabulated functions with interaction groups · 52ed82ec
  Peter Eastman authored Mar 28, 2016
  
  52ed82ec
04 Feb, 2016 1 commit
- Parallelized the code for downloading positions · 8d293880
  Peter Eastman authored Feb 04, 2016
  
  8d293880
21 Jan, 2016 1 commit
- Created property for disabling the PME stream · 3a0fa3b2
  Peter Eastman authored Jan 20, 2016
  
  3a0fa3b2
05 Nov, 2015 1 commit
- Fixed a race condition in energy computation · 5cb6ad03
  Peter Eastman authored Nov 05, 2015
  
  5cb6ad03
04 Nov, 2015 1 commit
- Finished implementing CompoundIntegrator · 5fa6fbc1
  Peter Eastman authored Nov 04, 2015
  
  5fa6fbc1
25 Oct, 2015 1 commit

Ensure context can be created when selecting CUDA device · 43926c46

Robert T. McGibbon authored Oct 23, 2015

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.

43926c46

21 Sep, 2015 1 commit
- Changes to generating nonbonded kernels · a47300e6
  Peter Eastman authored Sep 21, 2015
  
  a47300e6
18 Sep, 2015 1 commit
- Began CUDA implementation of periodicdistance() function for CustomExternalForce · 0b77ab91
  peastman authored Sep 18, 2015
  
  0b77ab91
03 Sep, 2015 1 commit
- Added NonbondedForce::getPMEParametersInContext() · 4c6d48ba
  peastman authored Sep 03, 2015
  
  4c6d48ba
02 Sep, 2015 1 commit
- Began CUDA version of CustomCentroidBondForce · 4d823b0a
  peastman authored Sep 02, 2015
  
  4d823b0a
27 Aug, 2015 1 commit
- Python 2/3 compatibility in single code base, plus python 3 testing on travis. · b7088b74
  peastman authored Aug 10, 2015
  
  b7088b74
10 Aug, 2015 1 commit
- Ensure neighbor list is large enough on first step · cb599d67
  Peter Eastman authored Aug 10, 2015
  
  cb599d67
29 Jul, 2015 1 commit
- Very minor optimizations · ef627240
  peastman authored Jul 29, 2015
  
  ef627240
16 Jul, 2015 3 commits
- CUDA CustomIntegrator supports if and while blocks · 506908ce
  peastman authored Jul 16, 2015
  
  506908ce
- Lots of cleanup to CUDA CustomIntegrator · 0360c770
  peastman authored Jul 16, 2015
  
  0360c770
- Began overhaul of CUDA CustomIntegrator in preparation for supporting flow control · 59c809c0
  peastman authored Jul 16, 2015
  
  59c809c0
07 Jul, 2015 1 commit
- CUDA allows different force groups to have different cutoffs · 85532a52
  Peter Eastman authored Jul 07, 2015
  
  85532a52
02 Jun, 2015 1 commit
- Cleaned up CUDA FFT code · 82069311
  Peter Eastman authored Jun 02, 2015
  
  82069311
30 May, 2015 1 commit
- Ported OpenCL FFT to CUDA · d3e91b15
  peastman authored May 29, 2015
  
  d3e91b15
15 May, 2015 1 commit
- Switched over to new Windows pthreads library · 6cf75568
  peastman authored May 15, 2015
  
  6cf75568
22 Apr, 2015 1 commit
- Fixed bug in CustomIntegrator · 4cf66cab
  Peter Eastman authored Apr 22, 2015
  
  4cf66cab
17 Mar, 2015 1 commit
- CUDA and OpenCL implementations of CMAPTorsionForce::updateParametersInContext() · 6dc2f9a8
  peastman authored Mar 16, 2015
  
  6dc2f9a8
16 Mar, 2015 1 commit
- Added updateParametersInContext() to CMAPTorsionForce, and implemented in reference platform · 474bbf7a
  peastman authored Mar 16, 2015
  
  474bbf7a
10 Mar, 2015 1 commit
- Fixed a performance regression in multi-GPU on CUDA · 9d3a655b
  peastman authored Mar 10, 2015
  
  9d3a655b
06 Mar, 2015 2 commits
- For multi-GPU, exceeding the size of the neighbor list causes the forces to be marked as invalid · bd666c27
  peastman authored Mar 06, 2015
  
  bd666c27
- Load balancing between GPUs forces the neighbor list to be rebuilt · 62f7c44c
  peastman authored Mar 06, 2015
  
  62f7c44c
13 Feb, 2015 1 commit
- Created plugin for CUDA runtime compiler · 7e29671a
  Peter Eastman authored Feb 13, 2015
  
  7e29671a
07 Jan, 2015 1 commit
- Fixed some methods that should have been marked const · 17c06daa
  Peter Eastman authored Jan 07, 2015
  
  17c06daa
05 Jan, 2015 1 commit
- Beginning of CUDA support for triclinic boxes · 61458a5f
  Peter Eastman authored Jan 05, 2015
  
  61458a5f
05 Nov, 2014 1 commit
- Further improvements to multi-GPU performance · 67a3b4c0
  peastman authored Nov 05, 2014
  
  67a3b4c0
21 Oct, 2014 2 commits
- Workaround for driver bugs that affect GTX 980 · 8e2fc4ea
  Peter Eastman authored Oct 21, 2014
  
  8e2fc4ea
- Optimizations to multi-GPU calculations · 390e0a6b
  root authored Oct 21, 2014
  
  390e0a6b