- Registered on: 04/21/2016
- Last connection: 05/31/2018
- 12:05 AM GROMACS Revision 85b8ed21: Add and enable PME OpenCL
- PME OpenCL kernels (spread, gather, solve) are added.
PmeGpuContext is taught to compile them. PME with OpenCL
- 12:31 PM GROMACS Revision bfeace90: Prepare pme-gpu-types.h for OpenCL
- DeviceBuffer fields are hidden from the future OpenCL
device compilation with a macro. Struct fields are reordered
- 05:53 PM GROMACS Revision dee2ffc7: Remove hardcoded warp_size == 32 assumption from PME GPU
- The PME OpenCL host code now queries the warp_size aka max
execution width from the device. The CUDA/unit test handli...
- 03:12 PM GROMACS Revision efc5b167: Move vectype_ops.clh into gpu_utils
- OpenCL compilation now uses gpu_utils as an include directory in source.
All the OpenCL sources are still installed i...
- 03:11 PM GROMACS Revision e553d9bc: Move the PME GPU kernel launch functions into pme-gpu-internal.cpp
- Change-Id: Iafd464a94aa12eb9c9743e3c3fae307208f2a3dd
- 08:44 AM GROMACS Revision abb0e3c6: Ensure PME with OpenCL does not attempt to pin
- Host-only memory pinning was designed with CUDA in mind, while OpenCL
requires managing both host and device memory b...
- 01:17 PM GROMACS Revision 1abb5f54: Compile most of PME GPU host code with OpenCL
- Change-Id: Ifca008ac44e345cc6ab87a3b7bd6d5694385a082
- 08:53 PM GROMACS Revision bfda3898: Merge pme.cu into pme-gpu-internal.cpp
- All the code is moved as is, only the includes change.
- 06:14 PM GROMACS Revision 29cb20e8: PME OpenCL 3D FFT transform is implemented, using clFFT
- This is not yet used but already compiled.
- 05:14 PM GROMACS Task #2537 (New): Simplify PME solve reduction
- Solve reduction (7 energy/virial components) in PME CUDA/OpenCL is written in a rather contrived way, and can be rewr...
Also available in: Atom