Commits · 222cc426fd5fa2254f17a854abdd123549dfea72 · Kaushik Kulkarni / viennacl-dev · GitLab

Aug 19, 2014
- NMF extended on CUDA,OpenMP. Docummentation of NMF finnished. Tests and... · 222cc426
  JurajKabzan authored Aug 19, 2014
  222cc426
Aug 17, 2014
- FFT extended on CUDA,OpenMP. Created multiple tests and tutorial. FFT... · b183be65
  JurajKabzan authored Aug 17, 2014
  b183be65
Aug 13, 2014
- SVD: Remove unused context variable · 46f90ebd
  Toby Smithe authored Aug 13, 2014
  
  46f90ebd
Aug 12, 2014
- AMG: Fixed invalid decrement of auxiliary iterator. · db903fb7
  Karl Rupp authored Aug 12, 2014
  db903fb7
- Context: Reverted commented check of mem_type for MEMORY_NOT_INITIALIZED · 08c2246a
  Karl Rupp authored Aug 12, 2014
  
  08c2246a
Aug 11, 2014
- ILU: context fixes · a8a007f7
  Toby Smithe authored Aug 11, 2014
  
  a8a007f7
- context: add VIENNACL_NO_CURRENT_CONTEXT to disable current_context with an... · f68e852e
  Toby Smithe authored Aug 11, 2014
  f68e852e
- OpenCL: Member functions for converting OpenCL device properties to string now public. · eb3f3602
  Karl Rupp authored Aug 11, 2014
  
  eb3f3602
- README: Updated instructions to create build folder and on how to build with Windows. · ae0d4a07
  Karl Rupp authored Aug 11, 2014
  
  ae0d4a07
Aug 10, 2014
- Some more context fixes (matrix_operations, svd) · a335d194
  Toby Smithe authored Aug 10, 2014
  
  a335d194
- Change cache_path to use std::string throughout, to fix weird string termination bugs in Python · 8c2a5bdf
  Toby Smithe authored Aug 10, 2014
  
  8c2a5bdf
Aug 09, 2014
- ocl::context: set _cache_path on construction (no need to wait for init), and add accessor methods · ce232f91
  Toby Smithe authored Aug 09, 2014
  
  ce232f91
Aug 08, 2014
- FFT: Create temporary objects with correct contexts · 7f7cd0b4
  Toby Smithe authored Aug 08, 2014
  
  7f7cd0b4
- BiCGStab: Added missing #include <numeric> · 0965e292
  Karl Rupp authored Aug 08, 2014
  
  0965e292
- NMF: eliminate a floating-point / conversion warning · ac92b722
  Toby Smithe authored Aug 08, 2014
  
  ac92b722
- BiCGStab: Added missing #include statements. · 6cb1124a
  Karl Rupp authored Aug 08, 2014
  6cb1124a
- More satisfactory patch for NMF 0-matrix bug: compute norms of initial W and... · 36352caa
  Toby Smithe authored Aug 08, 2014
  36352caa
- Merge branch 'karlrupp/feature-pipelined-bicgstab' · cc044483
  Karl Rupp authored Aug 08, 2014
  cc044483
- BiCGStab: Applied const-correctness in OpenCL and CUDA kernels. · b8c91799
  Karl Rupp authored Aug 08, 2014
  b8c91799
- BiCGStab: Added pipelined CUDA kernels. CUDA implementation now complete. · 39532998
  Karl Rupp authored Aug 08, 2014
  
  39532998
- BiCGStab: Added OpenCL implementation of pipelined kernels. · 314e816b
  Karl Rupp authored Aug 08, 2014
  
  314e816b
Aug 07, 2014
- NMF: Create objects using correct fixes, and don't stop after just one set of iterations · 8f3d97b9
  Toby Smithe authored Aug 07, 2014
  
  8f3d97b9
- BiCGStab: Completed implementation of host-based backend. · 52a9f1af
  Karl Rupp authored Aug 07, 2014
  
  52a9f1af
- BiCGStab: Added pipelined implementation for compressed_matrix with host-based backend. · 005e23b7
  Karl Rupp authored Aug 07, 2014
  005e23b7
Aug 06, 2014
- Merge branch 'karlrupp/feature-pipelined-cg' · 09e9eaa0
  Karl Rupp authored Aug 06, 2014
  09e9eaa0
- CG: Added CUDA kernels. Pipelined CG now complete. · 3aac6265
  Karl Rupp authored Aug 06, 2014
  
  3aac6265
- CG: Completed implementation of improved (pipelined) OpenCL kernels. · 4cd830a4
  Karl Rupp authored Aug 06, 2014
  
  4cd830a4
Aug 05, 2014
- CG: Added improved kernels for OpenCL when using compressed_matrix<> · 41c0c5a6
  Karl Rupp authored Aug 05, 2014
  41c0c5a6
Aug 04, 2014
- Tools: Added sparse matrix generation routine (FDM 2D) · 7237e71f
  Karl Rupp authored Aug 04, 2014
  7237e71f
- Pipelined CG: Added implementation for host-based execution. · 812e3918
  Karl Rupp authored Aug 04, 2014
  812e3918
- Merge branch 'wip-sliced-ell-matrix'. · 246370b0
  Karl Rupp authored Aug 04, 2014
  246370b0
- CUDA: Added option for specifying the CUDA arch through CMake. · d82ddd78
  Karl Rupp authored Aug 04, 2014
  d82ddd78
- sliced_ell_matrix: Support for Sliced ELL-matrix format (sliced_ell_matrix). · 1f9db9c9
  Karl Rupp authored Aug 04, 2014
  1f9db9c9
Aug 03, 2014
- Matrix: Added support for the operations A = trans(B), A += trans(B), A -= trans(B) · 0f8d1b6e
  Karl Rupp authored Aug 03, 2014
  0f8d1b6e
Aug 02, 2014
- VS 2012: Fixed compilation problem in compressed_compressed_matrix<> · ec267fcb
  Karl Rupp authored Aug 02, 2014
  ec267fcb
Aug 01, 2014
- Scheduler: Added missing implementations for element_pow() · 25076ce0
  Karl Rupp authored Aug 01, 2014
  
  25076ce0
- Tests: Reduced round-off errors in element_pow() test. · 2b8e01d7
  Karl Rupp authored Aug 01, 2014
  2b8e01d7
- VS2012: Fixed weird compilation problem with std::min() · 5c0ee23f
  Karl Rupp authored Aug 01, 2014
  5c0ee23f
- vector_operations: create temporary scalars with correct context · c3b474a1
  Toby Smithe authored Aug 01, 2014
  
  c3b474a1
Jul 27, 2014
- Device-Specific / GEMM : Now letting the compiler unroll the loop. · 5fae4e32
  Philippe Tillet authored Jul 27, 2014
  5fae4e32