Commits · 6deaa42f56d99662a52b7bea9955b12e70bfd72b · Kaushik Kulkarni / viennacl-dev · GitLab

Aug 17, 2014
- Tests: Rewrote the GEMM test from scratch. · 6deaa42f
  Philippe Tillet authored Aug 16, 2014
  6deaa42f
Aug 16, 2014
- OCL/Context : Fixed compilation problem when VIENNACL_DEBUG_ALL is defined... · 293e61b5
  Philippe Tillet authored Aug 16, 2014
  
  293e61b5
- Linalg/OpenCL : Normalized the arguments names, more blas-like. x,y,z for... · 305f1be8
  Philippe Tillet authored Aug 16, 2014
  305f1be8
- Linalg/OpenCL : Cleaned some dead code · 43018511
  Philippe Tillet authored Aug 16, 2014
  
  43018511
- Linalg/OpenCL : Fixed a couple of warnings in GEMM · 01962d8b
  Philippe Tillet authored Aug 16, 2014
  
  01962d8b
- OCL: ocl::program now stores a vector of smart pointers · 3e8b129c
  Philippe Tillet authored Aug 16, 2014
  
  3e8b129c
- OCL: ocl::context now stores a vector of smart pointers · 5089ce01
  Philippe Tillet authored Aug 16, 2014
  
  5089ce01
- Device-specific : Fixed some warnings · fc803979
  Philippe Tillet authored Aug 16, 2014
  
  fc803979
- Device-specific : Substantial improvements of the module. · 000eac96
  Philippe Tillet authored Aug 15, 2014
  
  000eac96
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