Commits · 32af3402d166c3782a799270c1e04e6d510257be · Kaushik Kulkarni / viennacl-dev · GitLab

Nov 13, 2012
- * Added matrix initializers (work for CPU and OpenCL, CUDA-testing required) · 32af3402
  Karl Rupp authored Nov 12, 2012
  32af3402
Nov 12, 2012
- Transfer RAM<->OpenCL<->CUDA<->RAM now implemented. · 24b5e039
  Karl Rupp authored Nov 12, 2012
  
  24b5e039
- Pimped ILU(0,T)-preconditioners, up to one order of magnitude faster with new... · 1b975b43
  Karl Rupp authored Nov 11, 2012
  1b975b43
Nov 11, 2012
- Reimplementation of ILU0 for compressed_matrix. Using the new multi-backend,... · 3a53cf71
  Karl Rupp authored Nov 11, 2012
  3a53cf71
- Sparse triangular solvers for compressed_matrix now working (CPU, OpenCL, CUDA). · 113e0c7f
  Karl Rupp authored Nov 11, 2012
  
  113e0c7f
- Sparse triangular solver for compressed_matrix now working on CPU. Moved in... · 49010db8
  Karl Rupp authored Nov 10, 2012
  49010db8
Nov 09, 2012
- Working on sparse triangular solvers. Performance still not great, but soon sufficient for ILU. · 98af43d5
  Karl Rupp authored Nov 09, 2012
  
  98af43d5
Nov 08, 2012
- Some progress with triangular solver for compressed_matrix: unit-lower-solve... · 81969744
  Karl Rupp authored Nov 08, 2012
  81969744
- * Corrected deprecated VIENNACL_HAVE_XYZ in examples · 43434a19
  Karl Rupp authored Nov 07, 2012
  43434a19
Nov 07, 2012
- Added VIENNACL_WITH_OPENMP guard to existing OpenMP stuff · 3b0b1dbe
  Karl Rupp authored Nov 07, 2012
  
  3b0b1dbe
- * Unified preprocessor defines for external toolkit: VIENNACL_WITH_XYZ. Old... · b35f793b
  Karl Rupp authored Nov 07, 2012
  b35f793b
Nov 06, 2012
- Completed CUDA backend by adding direct triangular solvers and LU factorization. All tests pass. · 5476f4ba
  Karl Rupp authored Nov 06, 2012
  
  5476f4ba
Nov 05, 2012
- * Reduced generic vector kernel (av, avbv, avbv_v) startup by 10-20 percent by packing arguments · 575300ac
  Karl Rupp authored Nov 05, 2012
  575300ac
- Fixed minor bugs in vector CUDA kernels (accidental uses of 'float' instead of... · a844270f
  Karl Rupp authored Nov 05, 2012
  a844270f
Nov 04, 2012
- Sparse matrices now working with CUDA. · b759bf80
  Karl Rupp authored Nov 04, 2012
  
  b759bf80
Nov 03, 2012
- Vector operations and matrix operations now work for CUDA. No matrix-matrix... · 8c815ad1
  Karl Rupp authored Nov 03, 2012
  8c815ad1
- Added CUDA skeleton for vector operations. Needs testing. · 4b9eb57d
  Karl Rupp authored Nov 02, 2012
  
  4b9eb57d
Nov 02, 2012
- Added memory handling and scalar operations for CUDA. · 6bc0e300
  Karl Rupp authored Nov 02, 2012
  
  6bc0e300
Nov 01, 2012
- Compilation with nvcc now works. Fixed all warnings and errors reported by nvcc. · d036b567
  Karl Rupp authored Nov 01, 2012
  
  d036b567
- Refurbished build system, now deals correctly with OpenCL-enabled and -disabled environments. · 9611f739
  Karl Rupp authored Nov 01, 2012
  
  9611f739
- * Added VIENNACL_HAVE_OPENCL flag for including OpenCL specific stuff. · 8c84096f
  Karl Rupp authored Oct 31, 2012
  8c84096f
Oct 30, 2012
- Cleaned up norm-kernels. Now using a single kernel for norm_1, norm_2 and... · 4baa4d13
  Karl Rupp authored Oct 30, 2012
  4baa4d13
- Purged old/unused kernels in auxiliary folder · 0ed4f0f7
  Karl Rupp authored Oct 30, 2012
  
  0ed4f0f7
- Fixed remaining issues with matrix(-range/slice) interactions. Removed... · 648d3c50
  Karl Rupp authored Oct 30, 2012
  648d3c50
- Added lots of operator overloads to matrices in order to handle them as... · ee28ab73
  Karl Rupp authored Oct 29, 2012
  ee28ab73
Oct 29, 2012
- Working on improved matrix-operations test. Adding missing operator overloads.... · 240613da
  Karl Rupp authored Oct 29, 2012
  240613da
Oct 28, 2012
- Added mixed precision CG solver. Further polishing required, yet the concept works. · a94c976a
  Karl Rupp authored Oct 28, 2012
  
  a94c976a
- Modified inner_prod() such that summation after multi-group reduction is... · 297def3b
  Karl Rupp authored Oct 28, 2012
  297def3b
Oct 27, 2012
- * Unified vector operations test, all combinations of vector, vector_range and... · 9950b23b
  Karl Rupp authored Oct 27, 2012
  9950b23b
Oct 26, 2012
- Refurbished operator overloads for vector: Moved member-overloads to global... · 9fab5e1b
  Karl Rupp authored Oct 25, 2012
  9fab5e1b
Oct 25, 2012
- Fixed 'problem' with sparse matrices: Row sums vanished, so test was using a... · 809c0f42
  Karl Rupp authored Oct 25, 2012
  809c0f42
- Significantly improved blas3-tests. All combinations of matrix/range/slice are... · ba99a9d7
  Karl Rupp authored Oct 25, 2012
  ba99a9d7
- Merged latest improvements to kernel generator by Philippe (previous commit).... · 7a8a3312
  Karl Rupp authored Oct 25, 2012
  7a8a3312
- Merge branch 'sf.net' into dev · 69d544b8
  Karl Rupp authored Oct 25, 2012
  
  69d544b8
- Two more fixes in order to compile cleanly on G++ 4.4. · be5062e7
  Karl Rupp authored Oct 25, 2012
  
  be5062e7
- ViennaCL now compiles cleanly with GCC 4.6.3 on -Wall -Wextra -pedantic... · 9a3ac85a
  Karl Rupp authored Oct 25, 2012
  9a3ac85a
- * Tried to fix the dense triangular solvers, yet my AMD driver is buggy... · a16b2f59
  Karl Rupp authored Oct 24, 2012
  a16b2f59
Oct 24, 2012
- Improved direct solver interface, now also accepts matrix_range and... · 21c4125a
  Karl Rupp authored Oct 24, 2012
  21c4125a
- Eliminated bug in handling transpose dense matrices on single-threaded CPU... · 4b6a641e
  Karl Rupp authored Oct 23, 2012
  4b6a641e
Oct 23, 2012
- Forgot to add new files in previous commit. · 13a96fc4
  Karl Rupp authored Oct 23, 2012
  
  13a96fc4