aboutsummaryrefslogtreecommitdiffhomepage
path: root/unsupported
Commit message (Collapse)AuthorAge
...
* Optimize TensorBlockCopyOpGravatar Eugene Zhulenev2018-09-27
|
* Revert code lost in mergeGravatar Eugene Zhulenev2018-09-27
|
* Merge with eigen/eigen defaultGravatar Eugene Zhulenev2018-09-27
|\
* | Remove explicit mkldnn support and redundant TensorContractionKernelBlockingGravatar Eugene Zhulenev2018-09-27
| |
* | Test mkldnn pack for doublesGravatar Eugene Zhulenev2018-09-26
| |
* | Conditionally add mkldnn testGravatar Eugene Zhulenev2018-09-26
| |
| * Remove "false &&" left over from test.Gravatar Rasmus Munk Larsen2018-09-26
| |
| * Parallelize tensor contraction over the inner dimension in cases where where ↵Gravatar Rasmus Munk Larsen2018-09-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | one or both of the outer dimensions (m and n) are small but k is large. This speeds up individual matmul microbenchmarks by up to 85%. Naming below is BM_Matmul_M_K_N_THREADS, measured on a 2-socket Intel Broadwell-based server. Benchmark Base (ns) New (ns) Improvement ------------------------------------------------------------------ BM_Matmul_1_80_13522_1 387457 396013 -2.2% BM_Matmul_1_80_13522_2 406487 230789 +43.2% BM_Matmul_1_80_13522_4 395821 123211 +68.9% BM_Matmul_1_80_13522_6 391625 97002 +75.2% BM_Matmul_1_80_13522_8 408986 113828 +72.2% BM_Matmul_1_80_13522_16 399988 67600 +83.1% BM_Matmul_1_80_13522_22 411546 60044 +85.4% BM_Matmul_1_80_13522_32 393528 57312 +85.4% BM_Matmul_1_80_13522_44 390047 63525 +83.7% BM_Matmul_1_80_13522_88 387876 63592 +83.6% BM_Matmul_1_1500_500_1 245359 248119 -1.1% BM_Matmul_1_1500_500_2 401833 143271 +64.3% BM_Matmul_1_1500_500_4 210519 100231 +52.4% BM_Matmul_1_1500_500_6 251582 86575 +65.6% BM_Matmul_1_1500_500_8 211499 80444 +62.0% BM_Matmul_3_250_512_1 70297 68551 +2.5% BM_Matmul_3_250_512_2 70141 52450 +25.2% BM_Matmul_3_250_512_4 67872 58204 +14.2% BM_Matmul_3_250_512_6 71378 63340 +11.3% BM_Matmul_3_250_512_8 69595 41652 +40.2% BM_Matmul_3_250_512_16 72055 42549 +40.9% BM_Matmul_3_250_512_22 70158 54023 +23.0% BM_Matmul_3_250_512_32 71541 56042 +21.7% BM_Matmul_3_250_512_44 71843 57019 +20.6% BM_Matmul_3_250_512_88 69951 54045 +22.7% BM_Matmul_3_1500_512_1 369328 374284 -1.4% BM_Matmul_3_1500_512_2 428656 223603 +47.8% BM_Matmul_3_1500_512_4 205599 139508 +32.1% BM_Matmul_3_1500_512_6 214278 139071 +35.1% BM_Matmul_3_1500_512_8 184149 142338 +22.7% BM_Matmul_3_1500_512_16 156462 156983 -0.3% BM_Matmul_3_1500_512_22 163905 158259 +3.4% BM_Matmul_3_1500_512_32 155314 157662 -1.5% BM_Matmul_3_1500_512_44 235434 158657 +32.6% BM_Matmul_3_1500_512_88 156779 160275 -2.2% BM_Matmul_1500_4_512_1 363358 349528 +3.8% BM_Matmul_1500_4_512_2 303134 263319 +13.1% BM_Matmul_1500_4_512_4 176208 130086 +26.2% BM_Matmul_1500_4_512_6 148026 115449 +22.0% BM_Matmul_1500_4_512_8 131656 98421 +25.2% BM_Matmul_1500_4_512_16 134011 82861 +38.2% BM_Matmul_1500_4_512_22 134950 85685 +36.5% BM_Matmul_1500_4_512_32 133165 90081 +32.4% BM_Matmul_1500_4_512_44 133203 90644 +32.0% BM_Matmul_1500_4_512_88 134106 100566 +25.0% BM_Matmul_4_1500_512_1 439243 435058 +1.0% BM_Matmul_4_1500_512_2 451830 257032 +43.1% BM_Matmul_4_1500_512_4 276434 164513 +40.5% BM_Matmul_4_1500_512_6 182542 144827 +20.7% BM_Matmul_4_1500_512_8 179411 166256 +7.3% BM_Matmul_4_1500_512_16 158101 155560 +1.6% BM_Matmul_4_1500_512_22 152435 155448 -1.9% BM_Matmul_4_1500_512_32 155150 149538 +3.6% BM_Matmul_4_1500_512_44 193842 149777 +22.7% BM_Matmul_4_1500_512_88 149544 154468 -3.3%
* | Support multiple contraction kernel types in TensorContractionThreadPoolGravatar Eugene Zhulenev2018-09-26
|/
* Don't deactivate BVH test for clang (probably, this was failing for very old ↵Gravatar Christoph Hertzberg2018-09-25
| | | | versions of clang)
* Fiw shadowing of last and allGravatar Gael Guennebaud2018-09-21
|
* Add missing plugins to DynamicSparseMatrix -- fix sparse_extra_3Gravatar Gael Guennebaud2018-09-21
|
* Collapsed revisionGravatar Eugene Zhulenev2018-09-20
| | | | | | * Split cxx11_tensor_executor test * Register test parts with EIGEN_SUFFIXES * Fix EIGEN_SUFFIXES in cxx11_tensor_executor test
* Merged eigen/eigen into defaultGravatar Rasmus Munk Larsen2018-09-20
|\
* | Initialize BlockIteratorState in a C++03 compatible way.Gravatar Rasmus Munk Larsen2018-09-20
| |
| * mergeGravatar Gael Guennebaud2018-09-20
| |\
| * | Fix MPrealSupportGravatar Gael Guennebaud2018-09-20
| | |
| * | Fix GPU support.Gravatar Gael Guennebaud2018-09-20
| | |
| * | Fix regression introduced by the previous fix for AVX512.Gravatar Gael Guennebaud2018-09-20
| | | | | | | | | | | | It brokes the complex-complex case on SSE.
| | * Workaround for spurious "array subscript is above array bounds" warnings ↵Gravatar Christoph Hertzberg2018-09-20
| | | | | | | | | | | | with g++4.x
| | * Avoid warning "suggest braces around initialization of subobject".Gravatar Christoph Hertzberg2018-09-20
| |/ |/| | | | | This test is not run in C++03 mode, so no compatibility is lost.
| * Fix gebp kernel for real+complex in case only reals are vectorized (e.g., ↵Gravatar Gael Guennebaud2018-09-20
|/ | | | | | AVX512). This commit also removes "half-packet" from data-mappers: it was not used and conceptually broken anyways.
* Cast to longer type.Gravatar Rasmus Munk Larsen2018-09-19
|
* Silence compiler warning.Gravatar Rasmus Munk Larsen2018-09-19
|
* Silence more compiler warnings.Gravatar Rasmus Munk Larsen2018-09-19
|
* Silence compiler warnings in ThreadPoolInterface.h.Gravatar Rasmus Munk Larsen2018-09-19
|
* Support static dimensions (aka IndexList) in Tensor::resize(...)Gravatar Eugene Zhulenev2018-09-18
|
* Enable DSizes type promotion with c++03 compilersGravatar Eugene Zhulenev2018-09-18
|
* Collapsed revisionGravatar Ravi Kiran2018-09-17
| | | | * Merged eigen/eigen into default
* Merged in ezhulenev/eigen-02 (pull request PR-498)Gravatar Rasmus Munk Larsen2018-09-17
|\ | | | | | | Add DSizes index type promotion
* \ Merged in ezhulenev/eigen-01 (pull request PR-497)Gravatar Rasmus Munk Larsen2018-09-17
|\ \ | | | | | | | | | Fix warnings in IndexList array_prod
| * | Replace deprecated Eigen::DenseIndex with Eigen::Index in TensorIndexListGravatar Eugene Zhulenev2018-09-17
| | |
* | | bug #1596: fix inclusion of Eigen's header within unsupported modules.Gravatar Gael Guennebaud2018-09-17
| | |
* | | typoGravatar Gael Guennebaud2018-09-16
| | |
| | * Add DSizes index type promotionGravatar Eugene Zhulenev2018-09-15
| |/ |/|
| * Fix warnings in IndexList array_prodGravatar Eugene Zhulenev2018-09-15
|/
* Fix weird error for building with g++-4.7 in C++03 mode.Gravatar Christoph Hertzberg2018-09-15
|
* Merged in ezhulenev/eigen/fix_dsizes (pull request PR-494)Gravatar Rasmus Munk Larsen2018-09-15
|\ | | | | | | Fix DSizes IndexList constructor
* \ Merged in ezhulenev/eigen/moar_eigen_fixes_3 (pull request PR-493)Gravatar Rasmus Munk Larsen2018-09-15
|\ \ | | | | | | | | | | | | | | | Const cast scalar pointer in TensorSlicingOp evaluator Approved-by: Sameer Agarwal <sameeragarwal@google.com>
| | * Fix DSizes IndexList constructorGravatar Eugene Zhulenev2018-09-14
| |/ |/|
* | Merged in ezhulenev/eigen/moar_eigen_fixes_1 (pull request PR-492)Gravatar Rasmus Munk Larsen2018-09-15
|\ \ | | | | | | | | | Explicitly construct tensor block dimensions from evaluator dimensions
| | * Const cast scalar pointer in TensorSlicingOp evaluatorGravatar Eugene Zhulenev2018-09-14
| |/ |/|
| * Explicitly construct tensor block dimensions from evaluator dimensionsGravatar Eugene Zhulenev2018-09-14
| |
* | Merged in chtz/eigen-maxsizevector (pull request PR-490)Gravatar Rasmus Munk Larsen2018-09-14
|\ \ | |/ |/| | | | | | | Let MaxSizeVector respect alignment of objects Approved-by: Rasmus Munk Larsen <rmlarsen@google.com>
* | Support reshaping with static shapes and dimensions conversion in tensor ↵Gravatar Eugene Zhulenev2018-09-14
| | | | | | | | broadcasting
| * bug #1598: Let MaxSizeVector respect alignment of objects and add a unit testGravatar Christoph Hertzberg2018-09-14
|/ | | | | Also revert 8b3d9ed081fc5d4870290649853b19cb5179546e
* Merged in rmlarsen/eigen3 (pull request PR-480)Gravatar Rasmus Munk Larsen2018-09-14
|\ | | | | | | Avoid compilation error in C++11 test when EIGEN_AVOID_STL_ARRAY is set.
| * Don't use bracket syntax in ctor.Gravatar Rasmus Munk Larsen2018-09-13
| |
* | Fix merge error.Gravatar Rasmus Munk Larsen2018-09-13
| |
* | Backed out changeset 01197e44527941c95f9a63e4f60ab3a989f12cbeGravatar Rasmus Munk Larsen2018-09-13
| |
generated by cgit v1.2.3 (git 2.39.1) at 2024-06-26 00:32:42 +0000