Commits · 92dc4a4720d74e31eeb8cb2e7e0ee724c7f057e1 · TNL / tnl-dev · GitLab

Snippets Groups Projects

This is an archived project. Repository and other project resources are read-only.

Sep 02, 2019
- Renaming PrefixSum to Scan. · 92dc4a47
  Tomáš Oberhuber authored 5 years ago
  
  92dc4a47
- Fixed building documentation with code and output snippets · 98177777
  Jakub Klinkovský authored 5 years ago
  
  98177777
- Added documenation of some vector functions plus reduction tutorial fixes. · 10d646e6
  Tomáš Oberhuber authored 5 years ago
  
  10d646e6
- Writting tutorial on parallel reduction. · 2d0b5ea5
  Tomáš Oberhuber authored 5 years ago
  
  2d0b5ea5
Aug 27, 2019
- Avoiding compiler warnings for builds without CUDA · 7390a03b
  Jakub Klinkovský authored 5 years ago
  
  7390a03b
- Fixed previous commit · 74c653f1
  Jakub Klinkovský authored 5 years ago
  
  74c653f1
- Fixed previous commit · d27c1928
  Jakub Klinkovský authored 5 years ago
  
  d27c1928
Aug 26, 2019
- Removed useless using namespace declaration from DistributedGrid tests · 449aa717
  Jakub Klinkovský authored 5 years ago
  449aa717
Aug 25, 2019
- Removed useless includes and specialization of SubdomainOverlapsGetter · 688e3bd1
  Jakub Klinkovský authored 5 years ago
  
  688e3bd1
- Avoiding compiler warnings · 238c404b
  Jakub Klinkovský authored 5 years ago
  
  238c404b
Aug 24, 2019
- Avoiding nvcc warnings in NDArray internals · a4157051
  Jakub Klinkovský authored 5 years ago
  
  a4157051
- Avoiding compiler warnings · 8253355f
  Jakub Klinkovský authored 5 years ago
  
  8253355f
Aug 17, 2019
- Implemented distributed prefix-sum · d13a2d18
  Jakub Klinkovský authored 5 years ago
  d13a2d18
- PrefixSum: separate first and second phase for OpenMP implementation and... · 174ad5fd
  Jakub Klinkovský authored 5 years ago
  174ad5fd
- CUDA prefix-sum: separated the implementation of the first and second phase · 2c40015f
  Jakub Klinkovský authored 5 years ago
  
  2c40015f
- CUDA prefix-sum: moved gridShift from the first phase to the second phase · ac2ee07e
  Jakub Klinkovský authored 5 years ago
  
  ac2ee07e
- Replaced static member variables in CudaPrefixSumKernelLauncher with static getters · 1fe62640
  Jakub Klinkovský authored 5 years ago
  
  1fe62640
- Added default stream synchronizations after kernel launches in CudaPrefixSumKernel.h · af6d1d6b
  Jakub Klinkovský authored 5 years ago
  
  af6d1d6b
- Removed volatile reduction from PrefixSum and updated the normal reduction operation · 8d0d2638
  Jakub Klinkovský authored 5 years ago
  8d0d2638
- Added prefix-sum to BLAS benchmarks · 27631930
  Jakub Klinkovský authored 5 years ago
  
  27631930
- Implemented parallel prefix-sum with OpenMP · 7cc55dee
  Jakub Klinkovský authored 5 years ago
  7cc55dee
- Benchmarks: compute sample standard deviation of the measured computation times · 2bea9311
  Jakub Klinkovský authored 5 years ago
  
  2bea9311
- Removed timing parameter from benchmarks · e6e6cf46
  Jakub Klinkovský authored 5 years ago
  e6e6cf46
- Benchmarks: added scalar multiplication with BLAS · 232be124
  Jakub Klinkovský authored 5 years ago
  
  232be124
- Optimized OpenMP thread counts for reduction and multireduction · 505d0b68
  Jakub Klinkovský authored 5 years ago
  
  505d0b68
- Ugly workaround for nvcc's stupid modification of `new` expressions · 32c69a11
  Jakub Klinkovský authored 5 years ago
  
  32c69a11
- Replaced custom lambda functions with instances of STL types where possible · 0a57393f
  Jakub Klinkovský authored 5 years ago
  
  0a57393f
- Changed reduction operation to use functions with `return a + b` instead of `a += b` · e20a0930
  Jakub Klinkovský authored 5 years ago
  e20a0930
- Removed VectorOperations class which is now useless · d0fc1bb7
  Jakub Klinkovský authored 5 years ago
  d0fc1bb7
- Removed ReductionOperations.h · 1777e488
  Jakub Klinkovský authored 5 years ago
  
  1777e488
- Removed volatile reduction completely · 13b89a71
  Jakub Klinkovský authored 5 years ago
  
  13b89a71
- Found a way to avoid using volatile in CUDA reduction: __syncwarp() · cbc2fff9
  Jakub Klinkovský authored 5 years ago
  cbc2fff9
- Rewritten multireduction with lambda functions · b74a24d2
  Jakub Klinkovský authored 5 years ago
  
  b74a24d2
- Style changes in the code for reduction · e470040a
  Jakub Klinkovský authored 5 years ago
  
  e470040a
- Removed reduction and multireduction declarations for MIC · 1a82b047
  Jakub Klinkovský authored 5 years ago
  1a82b047
- Tests: forced using a unique file name in each test · a3ba2469
  Jakub Klinkovský authored 5 years ago
  a3ba2469
- Disabled long double in tnl-diff, tnl-init, tnl-lattice-init and tnl-view · 307b6650
  Jakub Klinkovský authored 5 years ago
  307b6650
Aug 16, 2019
- Cleaned up communicators in tnl-init · 9cda013b
  Jakub Klinkovský authored 5 years ago
  
  9cda013b
- Fixed move-constructor in Array · 569ccc9a
  Jakub Klinkovský authored 5 years ago
  
  569ccc9a
Aug 14, 2019
- DistributedNDArray: added methods getLocalIndexer, getLocalView,... · 05634323
  Jakub Klinkovský authored 5 years ago
  05634323