Commits · 9434c6e860c22f21ef1eedf04da78b6eda8e568a · tsoc / openmm

29 Jan, 2014 2 commits
- Created OpenCL and CUDA implementations of Continuous2DFunction · 9434c6e8
  peastman authored Jan 29, 2014
  
  9434c6e8
- Created reference implementation of Continuous2DFunction · 4cfeb429
  peastman authored Jan 28, 2014
  
  4cfeb429
27 Jan, 2014 1 commit

Tabulated function parameters are hardcoded in the kernel instead of being... · 0a1a011d

peastman authored Jan 27, 2014

Tabulated function parameters are hardcoded in the kernel instead of being stored in an array.  This makes the code simpler and may help performance slightly.

0a1a011d

25 Jan, 2014 1 commit
- Created Discrete2DFunction and Discrete3DFunction · a773952e
  peastman authored Jan 24, 2014
  
  a773952e
24 Jan, 2014 2 commits
- Created Discrete1DFunction · 56e36449
  peastman authored Jan 24, 2014
  
  56e36449
- Converted CUDA and OpenCL platforms to process tabulated functions with the TabulatedFunction API · 81f26683
  peastman authored Jan 23, 2014
  
  81f26683
23 Jan, 2014 1 commit
- Converted reference platform to process tabulated functions with the TabulatedFunction API · 98640b63
  peastman authored Jan 23, 2014
  
  98640b63
22 Jan, 2014 3 commits
- Fixed a compilation error · 77a54244
  peastman authored Jan 22, 2014
  
  77a54244
- Parallelized the initialization of positions and forces in the CPU platform · 146b66bd
  peastman authored Jan 22, 2014
  
  146b66bd
- Parallelized the computation of torsions with CPU platform · 543d51d7
  peastman authored Jan 22, 2014
  
  543d51d7
16 Jan, 2014 1 commit
- Fixed a test case that occasionally failed · feb0368f
  Peter Eastman authored Jan 15, 2014
  
  feb0368f
15 Jan, 2014 1 commit
- Test cases detect if CPU is not supported and exit rather than crashing · 9966f134
  Peter Eastman authored Jan 14, 2014
  
  9966f134
14 Jan, 2014 1 commit
- Detect Visual Studio in a way that works correctly with NMake files · 19dd114e
  peastman authored Jan 13, 2014
  
  19dd114e
11 Jan, 2014 2 commits
- Same bug fix as the previous commit, but for CUDA · 1e857c31
  peastman authored Jan 10, 2014
  
  1e857c31
- Fixed a bug in OpenCL implementation of CustomGBForce · 44e65cb1
  peastman authored Jan 10, 2014
  
  44e65cb1
09 Jan, 2014 3 commits
- Fixed additional errors on Windows · 76d592d6
  Peter Eastman authored Jan 09, 2014
  
  76d592d6
- Fixed a bug that could potentially cause LJ energy to be nan · 5c983283
  peastman authored Jan 09, 2014
  
  5c983283
- Optimizations to CPU GBSA · cf52e69f
  peastman authored Jan 08, 2014
  
  cf52e69f
08 Jan, 2014 1 commit
- Fixed lots of compilation errors on Windows · 72e81e04
  Peter Eastman authored Jan 08, 2014
  
  72e81e04
07 Jan, 2014 1 commit
- Very minor optimization · 06db7ac4
  peastman authored Jan 07, 2014
  
  06db7ac4
06 Jan, 2014 2 commits
- Improved logic for selecting a platform to better deal with exceptions in context initialization · d91e3f19
  peastman authored Jan 06, 2014
  
  d91e3f19
- Fixed some bugs identified by Valgrind · fc1d0b5c
  peastman authored Jan 06, 2014
  
  fc1d0b5c
21 Dec, 2013 2 commits
- Fixed quotes in command line when using CudaHostCompiler option · cc11a449
  peastman authored Dec 20, 2013
  
  cc11a449
- Parallelized summing the forces from different threads · 8bdf0829
  peastman authored Dec 20, 2013
  
  8bdf0829
19 Dec, 2013 3 commits
- Parallelized LangevinIntegrator for CPU platform · ddf1216d
  peastman authored Dec 19, 2013
  
  ddf1216d
- Fixed bug in detecting if AVX is supported · 6cde1860
  peastman authored Dec 19, 2013
  
  6cde1860
- Automatically select whether to use the SSE or AVX version of CpuNonbondedForce · d4af8747
  peastman authored Dec 18, 2013
  
  d4af8747
18 Dec, 2013 1 commit
- Improved vectorization of CpuNonbondedForce · 5cd23acb
  peastman authored Dec 18, 2013
  
  5cd23acb
17 Dec, 2013 1 commit
- Created AVX implementation of CPU nonbonded force · e22808f3
  peastman authored Dec 16, 2013
  
  e22808f3
13 Dec, 2013 2 commits
- Added option for setting CUDA host compiler · b989fd9b
  peastman authored Dec 13, 2013
  
  b989fd9b
- Reference implementation of AndersenThermostat did not handle massless particles correctly · 8d062ebb
  peastman authored Dec 12, 2013
  
  8d062ebb
10 Dec, 2013 1 commit
- Minor optimizations to CPU platform · 9ec4f597
  peastman authored Dec 10, 2013
  
  9ec4f597
09 Dec, 2013 1 commit
- Parallelized SETTLE on CPU platform · 06748808
  peastman authored Dec 09, 2013
  
  06748808
06 Dec, 2013 2 commits
- Major simplification to constraint handling in reference platform · 4d25f4e7
  peastman authored Dec 06, 2013
  
  4d25f4e7
- Further optimizations to CPU platform · 99f3a8a2
  peastman authored Dec 06, 2013
  
  99f3a8a2
05 Dec, 2013 2 commits
- Minor optimizations to CPU platform · 2882737d
  peastman authored Dec 05, 2013
  
  2882737d
- More fixes for problems on OS X 10.9 · 2cd817b4
  peastman authored Dec 04, 2013
  
  2cd817b4
04 Dec, 2013 2 commits
- Enforce memory alignment to improve performance of vector operations. Also... · b1be68d8
  peastman authored Dec 04, 2013
```
Enforce memory alignment to improve performance of vector operations.  Also fixed bugs in an earlier optimization.
```
  b1be68d8
- Optimizations to CPU nonbonded forces: better load balancing between threads,... · be04ec57
  peastman authored Dec 04, 2013
```
Optimizations to CPU nonbonded forces: better load balancing between threads, use linear splines instead of cubic
```
  be04ec57
03 Dec, 2013 1 commit
- Optimization to building neighbor list on CPU platform · 56cf0fde
  peastman authored Dec 03, 2013
  
  56cf0fde