Commit message (Collapse) | Author | Age | |
---|---|---|---|
* | Added SYCL include in Tensor. | 2016-10-20 | |
| | |||
* | Fixing the typo regarding missing #if needed for proper handling of ↵ | 2016-10-16 | |
| | | | | exceptions in Eigen/Core. | ||
* | Fixing the code indentation in the TensorReduction.h file. | 2016-10-14 | |
| | |||
* | Merged ComputeCpp to default. | 2016-10-14 | |
|\ | |||
| * | Applyiing Benoit's comment to return the missing line back in Eigen/Core | 2016-10-14 | |
| | | |||
* | | Merged ComputeCpp into default. | 2016-10-14 | |
|\| | |||
| * | Reducing the code by generalising sycl backend functions/structs. | 2016-10-14 | |
| | | |||
* | | Merged eigen/eigen into default | 2016-10-12 | |
|\ \ | |||
| * | | Remove double ;; | 2016-10-12 | |
| | | | |||
| * | | Fix SPQR for rectangular matrices | 2016-10-12 | |
| | | | |||
| * | | Fix outer-stride. | 2016-10-12 | |
| | | | |||
| * | | Merged in rmlarsen/eigen (pull request PR-230) | 2016-10-12 | |
| |\ \ | | | | | | | | | | | | | Fix a bug in psqrt for SSE and AVX when EIGEN_FAST_MATH=1 | ||
| | * | | Fix copy-paste error: Must use _mm256_cmp_ps for AVX. | 2016-10-12 | |
| | | | | |||
| * | | | bug #1325: fix compilation on NEON with clang | 2016-10-12 | |
| | | | | |||
| * | | | Manually define int16_t and uint16_t when compiling with Visual Studio | 2016-10-08 | |
| | | | | |||
| * | | | Reenabled the use of variadic templates on tegra x1 provides that the latest ↵ | 2016-10-08 | |
| | | | | | | | | | | | | | | | | version (i.e. JetPack 2.3) is used. | ||
| * | | | Cleaned up a regression test | 2016-10-08 | |
| | | | | |||
* | | | | Merge the content of the ComputeCpp branch into the default branch | 2016-10-07 | |
|\ \ \ \ | | |_|/ | |/| | | |||
| | * | | Remove static qualifier of free-functions (inline is enough and this helps ↵ | 2016-10-07 | |
| | | | | | | | | | | | | | | | | ICC to find the right overload) | ||
| | * | | Merged in rryan/eigen/tensorfunctors (pull request PR-233) | 2016-10-06 | |
| |/| | |/| | | | | | | | | | | | Fully support complex types in SumReducer and MeanReducer when building for CUDA by using scalar_sum_op and scalar_product_op instead of operator+ and operator*. | ||
| | * | | Add a test that GPU complex product reductions match CPU reductions. | 2016-10-06 | |
| | | | | |||
| | * | | Fully support complex types in SumReducer and MeanReducer when building for ↵ | 2016-10-06 | |
| | | | | | | | | | | | | | | | | CUDA by using scalar_sum_op and scalar_product_op instead of operator+ and operator*. | ||
* | | | | Added missing AVX intrinsics for fp16: in particular, implemented predux ↵ | 2016-10-06 | |
| |/ / |/| | | | | | | | | which is required by the matrix-vector code. | ||
* | | | Fix compilation of qr.inverse() for column and full pivoting variants. | 2016-10-06 | |
| | | | |||
| * | | Fixed a couple of compilation warnings | 2016-10-05 | |
| | | | |||
| * | | Pull the latest updates from trunk | 2016-10-05 | |
| |\ \ | |||
| * | | | Fixed compilation warnings | 2016-10-05 | |
| | | | | |||
| * | | | Fixed compilation warning | 2016-10-05 | |
| | | | | |||
* | | | | Increased the robustness of the reduction tests on fp16 | 2016-10-05 | |
| | | | | |||
* | | | | Increase the tolerance to numerical noise. | 2016-10-05 | |
| | | | | |||
* | | | | ::rand() returns a signed integer on win32 | 2016-10-05 | |
| | | | | |||
* | | | | Fixed a typo that impacts windows builds | 2016-10-05 | |
| |/ / |/| | | |||
* | | | Silenced compilation warning | 2016-10-04 | |
| | | | |||
* | | | Properly characterize the CUDA packet primitives for fp16 as device only | 2016-10-04 | |
| | | | |||
| | * | Update comment for fast sqrt. | 2016-10-04 | |
| | | | |||
| | * | Update comment for fast sqrt. | 2016-10-04 | |
| | | | |||
| | * | Fix a bug in the implementation of Carmack's fast sqrt algorithm in Eigen ↵ | 2016-10-04 | |
| |/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | (enabled by EIGEN_FAST_MATH), which causes the vectorized parts of the computation to return -0.0 instead of NaN for negative arguments. Benchmark speed in Giga-sqrts/s Intel(R) Xeon(R) CPU E5-1650 v3 @ 3.50GHz ----------------------------------------- SSE AVX Fast=1 2.529G 4.380G Fast=0 1.944G 1.898G Fast=1 fixed 2.214G 3.739G This table illustrates the worst case in terms speed impact: It was measured by repeatedly computing the sqrt of an n=4096 float vector that fits in L1 cache. For large vectors the operation becomes memory bound and the differences between the different versions almost negligible. | ||
* | | Cleanup the cuda executor code. | 2016-10-04 | |
| | | |||
* | | Cleaned up the random number generation code. | 2016-10-04 | |
| | | |||
* | | Use explicit type casting to generate packets of zeros. | 2016-10-04 | |
| | | |||
* | | Improved support for compiling CUDA code with clang as the host compiler | 2016-10-03 | |
| | | |||
* | | Added support for constand std::complex numbers on GPU | 2016-10-03 | |
| | | |||
* | | bug #1317: fix performance regression with some Block expressions and clang ↵ | 2016-10-01 | |
| | | | | | | | | | | | | by helping it to remove dead code. The trick is to get rid of the nested expression in the evaluator by copying only the required information (here, the strides). | ||
* | | bug #1310: workaround a compilation regression from 3.2 regarding triangular ↵ | 2016-09-30 | |
| | | | | | | | | * homogeneous | ||
| * | Renamed the SYCL tests to follow the standard naming convention. | 2016-09-30 | |
| | | |||
* | | Fix angle range | 2016-09-30 | |
| | | |||
* | | Remove std:: prefix | 2016-09-30 | |
| | | |||
* | | bug #1312: Quaternion to AxisAngle conversion now ensures the angle will be ↵ | 2016-09-29 | |
| | | | | | | | | in the range [-pi,pi]. This also increases accuracy when q.w is negative. | ||
* | | bug #1308: fix compilation of some small products involving nullary-expressions. | 2016-09-29 | |
| | | |||
* | | Updated the list of warnings to reflect the new message ids introduced in ↵ | 2016-09-28 | |
| | | | | | | | | cuda 8.0 |