Commits · 9ab32c50de91b13266998924c62145a073c74f4a · tsoc / openmm

12 Sep, 2011 1 commit
- Reduced the number of kernel executions to clear buffers · 9ab32c50
  Peter Eastman authored Sep 12, 2011
  
  9ab32c50
09 Sep, 2011 3 commits
- Fixed error on GPUs with only 16K local memory · 61c50862
  Peter Eastman authored Sep 09, 2011
  
  61c50862
- Minor optimizations to GBSA · 91f3379b
  Peter Eastman authored Sep 09, 2011
  
  91f3379b
- Minor performance optimizations · d2a5b3bb
  Peter Eastman authored Sep 09, 2011
  
  d2a5b3bb
08 Sep, 2011 1 commit
- Optimizations to PME reciprocal space calculation · 592dc5a9
  Peter Eastman authored Sep 08, 2011
  
  592dc5a9
07 Sep, 2011 1 commit
- Optimizations to FFT · c08d8a53
  Peter Eastman authored Sep 07, 2011
  
  c08d8a53
02 Sep, 2011 3 commits
- Additional minor optimizations to PME direct space computation · e85c741b
  Peter Eastman authored Sep 02, 2011
  
  e85c741b
- Using fast approximation for erfc instead of tabulated values · 0e5d3fb1
  Peter Eastman authored Sep 02, 2011
  
  0e5d3fb1
- On Fermi, store forces as 64 bit fixed point and update with atomic... · d9029f61
  Peter Eastman authored Sep 02, 2011
```
On Fermi, store forces as 64 bit fixed point and update with atomic operations.  Also various other minor optimizations.
```
  d9029f61
01 Sep, 2011 1 commit
- Added CL_CALLBACK to errorCallback() for Windows compile · 8ade176c
  Mark Friedrichs authored Sep 01, 2011
  
  8ade176c
26 Aug, 2011 1 commit
- Added callback function to print out internal OpenCL errors · e69b1cca
  Peter Eastman authored Aug 26, 2011
  
  e69b1cca
22 Aug, 2011 1 commit
- Fixed another case where the exclusions for CustomGBForce could be handled incorrectly · 84a14401
  Peter Eastman authored Aug 22, 2011
  
  84a14401
18 Aug, 2011 4 commits
- Fixed bug when CustomGBForce was told to use exclusions, but didn't have any... · e274d455
  Peter Eastman authored Aug 18, 2011
```
Fixed bug when CustomGBForce was told to use exclusions, but didn't have any exclusions.  (See bug 1519.)
```
  e274d455
- Change suggested by Tony Tye to simplify debugging by including all... · 6ab42996
  Peter Eastman authored Aug 18, 2011
```
Change suggested by Tony Tye to simplify debugging by including all compilation options in the kernel source
```
  6ab42996
- Fixed bug in clearAutoclearBuffers() · 696d90d5
  Peter Eastman authored Aug 18, 2011
  
  696d90d5
- Fixed bugs in OpenCL RPMD · 7f1da383
  Peter Eastman authored Aug 18, 2011
  
  7f1da383
13 Aug, 2011 1 commit
- Created OpenCL implementation of RPMD · a713aabf
  Peter Eastman authored Aug 13, 2011
  
  a713aabf
09 Aug, 2011 1 commit
- Fixed bugs that caused constraints to fail when some were implemented with... · 192b5c6b
  Peter Eastman authored Aug 09, 2011
```
Fixed bugs that caused constraints to fail when some were implemented with CCMA and some with a different method
```
  192b5c6b
22 Jul, 2011 2 commits
- Bug fix in AMD code · 53d1c90e
  Peter Eastman authored Jul 22, 2011
  
  53d1c90e
- addition of OPENMMCUDA_EXPORT for Windows compile · 97c767f8
  Mark Friedrichs authored Jul 22, 2011
  
  97c767f8
18 Jul, 2011 2 commits
- Fixed race condition in parallel execution where positions could get overwritten with forces · f794f818
  Peter Eastman authored Jul 18, 2011
  
  f794f818
- Reduced number of force buffers on AMD · 952757a8
  Peter Eastman authored Jul 18, 2011
  
  952757a8
16 Jul, 2011 1 commit

Only use the atomic version of charge spreading on compute level 2.0 or later,... · 95b02c8e

Peter Eastman authored Jul 16, 2011

Only use the atomic version of charge spreading on compute level 2.0 or later, since it is slower than the other version on earlier GPUs.

95b02c8e

15 Jul, 2011 3 commits
- Fixed compilation error on Linux · 25881dc0
  Peter Eastman authored Jul 15, 2011
  
  25881dc0
- Upgraded to newer version of cl.hpp · eeb3e32c
  Peter Eastman authored Jul 15, 2011
  
  eeb3e32c
- Fixed bug in identifying shake clusters · c14b0b8e
  Peter Eastman authored Jul 15, 2011
  
  c14b0b8e
07 Jul, 2011 3 commits
- Switched back to always using the bundled cl.hpp, since the previous code... · f9d0115f
  Peter Eastman authored Jul 07, 2011
```
Switched back to always using the bundled cl.hpp, since the previous code didn't work on some OpenCL implementations
```
  f9d0115f
- Added lots of test cases for multi-GPU computation · aea33abb
  Peter Eastman authored Jul 07, 2011
  
  aea33abb
- Optimized CCMA · d98bd6b0
  Peter Eastman authored Jul 07, 2011
  
  d98bd6b0
05 Jul, 2011 2 commits
- Fixed error when using Nvidia GPU on Apple's OpenCL · ac6ab5b7
  Peter Eastman authored Jul 05, 2011
  
  ac6ab5b7
- Implemented faster version of charge spreading on GPUs that support 64 bit atomics · ec3ee586
  Peter Eastman authored Jul 05, 2011
  
  ec3ee586
13 Jun, 2011 1 commit
- Fixed bug where angles at 180 degrees would sometimes produce nans · 38839771
  Peter Eastman authored Jun 13, 2011
  
  38839771
27 May, 2011 1 commit
- Optimization to improve cache coherency in charge spreading kernel · 839ab51e
  Peter Eastman authored May 27, 2011
  
  839ab51e
12 May, 2011 1 commit
- Optimized communication between GPUs by using pinned memory · 72dc8dfe
  Peter Eastman authored May 12, 2011
  
  72dc8dfe
09 May, 2011 2 commits
- Restructured the use of force buffers in a new way that hopefully really works everywhere. · 7943a339
  Peter Eastman authored May 09, 2011
  
  7943a339
- Optimizations on AMD suggested by Tony Tye · 13ef0ee8
  Peter Eastman authored May 09, 2011
  
  13ef0ee8
04 May, 2011 4 commits
- Use the preinstalled cl.hpp if it's available · e68471ec
  Peter Eastman authored May 04, 2011
  
  e68471ec
- Very minor optimizations · a7b68da3
  Peter Eastman authored May 04, 2011
  
  a7b68da3
- Fix to the previous checkin (which broke load balancing) · c56b4801
  Peter Eastman authored May 04, 2011
  
  c56b4801
- Minor optimizations to data transfer between GPUs · 5a9897bd
  Peter Eastman authored May 04, 2011
  
  5a9897bd