diff options
author | Vamsi Sripathi <vamsi.sripathi@intel.com> | 2018-05-23 14:02:05 -0700 |
---|---|---|
committer | Vamsi Sripathi <vamsi.sripathi@intel.com> | 2018-05-23 14:02:05 -0700 |
commit | 6293ad3f392a7b97ebb9f9f874682505c1391f2d (patch) | |
tree | 27e8306ff9ee047f8fce7bcd853fca06294b9783 /unsupported/test/cxx11_tensor_broadcasting.cpp | |
parent | 7134fa7a2eea469f35ea12899e693a633b5b42e5 (diff) |
Performance improvements to tensor broadcast operation
1. Added new packet functions using SIMD for NByOne, OneByN cases
2. Modified existing packet functions to reduce index calculations when input stride is non-SIMD
3. Added 4 test cases to cover the new packet functions
Diffstat (limited to 'unsupported/test/cxx11_tensor_broadcasting.cpp')
-rw-r--r-- | unsupported/test/cxx11_tensor_broadcasting.cpp | 62 |
1 files changed, 62 insertions, 0 deletions
diff --git a/unsupported/test/cxx11_tensor_broadcasting.cpp b/unsupported/test/cxx11_tensor_broadcasting.cpp index 5c0ea5889..a9d268ea6 100644 --- a/unsupported/test/cxx11_tensor_broadcasting.cpp +++ b/unsupported/test/cxx11_tensor_broadcasting.cpp @@ -180,6 +180,64 @@ static void test_fixed_size_broadcasting() #endif } +template <int DataLayout> +static void test_simple_broadcasting_one_by_n() +{ + Tensor<float, 4, DataLayout> tensor(1,13,5,7); + tensor.setRandom(); + array<ptrdiff_t, 4> broadcasts; + broadcasts[0] = 9; + broadcasts[1] = 1; + broadcasts[2] = 1; + broadcasts[3] = 1; + Tensor<float, 4, DataLayout> broadcast; + broadcast = tensor.broadcast(broadcasts); + + VERIFY_IS_EQUAL(broadcast.dimension(0), 9); + VERIFY_IS_EQUAL(broadcast.dimension(1), 13); + VERIFY_IS_EQUAL(broadcast.dimension(2), 5); + VERIFY_IS_EQUAL(broadcast.dimension(3), 7); + + for (int i = 0; i < 9; ++i) { + for (int j = 0; j < 13; ++j) { + for (int k = 0; k < 5; ++k) { + for (int l = 0; l < 7; ++l) { + VERIFY_IS_EQUAL(tensor(i%1,j%13,k%5,l%7), broadcast(i,j,k,l)); + } + } + } + } +} + +template <int DataLayout> +static void test_simple_broadcasting_n_by_one() +{ + Tensor<float, 4, DataLayout> tensor(7,3,5,1); + tensor.setRandom(); + array<ptrdiff_t, 4> broadcasts; + broadcasts[0] = 1; + broadcasts[1] = 1; + broadcasts[2] = 1; + broadcasts[3] = 19; + Tensor<float, 4, DataLayout> broadcast; + broadcast = tensor.broadcast(broadcasts); + + VERIFY_IS_EQUAL(broadcast.dimension(0), 7); + VERIFY_IS_EQUAL(broadcast.dimension(1), 3); + VERIFY_IS_EQUAL(broadcast.dimension(2), 5); + VERIFY_IS_EQUAL(broadcast.dimension(3), 19); + + for (int i = 0; i < 7; ++i) { + for (int j = 0; j < 3; ++j) { + for (int k = 0; k < 5; ++k) { + for (int l = 0; l < 19; ++l) { + VERIFY_IS_EQUAL(tensor(i%7,j%3,k%5,l%1), broadcast(i,j,k,l)); + } + } + } + } +} + void test_cxx11_tensor_broadcasting() { @@ -191,4 +249,8 @@ void test_cxx11_tensor_broadcasting() CALL_SUBTEST(test_static_broadcasting<RowMajor>()); CALL_SUBTEST(test_fixed_size_broadcasting<ColMajor>()); CALL_SUBTEST(test_fixed_size_broadcasting<RowMajor>()); + CALL_SUBTEST(test_simple_broadcasting_one_by_n<RowMajor>()); + CALL_SUBTEST(test_simple_broadcasting_n_by_one<RowMajor>()); + CALL_SUBTEST(test_simple_broadcasting_one_by_n<ColMajor>()); + CALL_SUBTEST(test_simple_broadcasting_n_by_one<ColMajor>()); } |