Merge branch 'amd-develop-0605' into amd-master

17cf8179 · Jun Liu · 6b9a4bd5 · e4112de7 · 17cf8179 · 17cf8179
Commit 17cf8179 authored Jun 05, 2024 by Jun Liu
20 changed files
--- a/example/21_gemm_layernorm/gemm_layernorm_xdl_naive_fp16.cpp
+++ b/example/21_gemm_layernorm/gemm_layernorm_xdl_naive_fp16.cpp
--- a/example/26_contraction/README.md
+++ b/example/26_contraction/README.md
@@ -7,14 +7,3 @@
 #arg3: time kernel (0=no, 1=yes)
 ./bin/example_contraction_bilinear_xdl_fp32 1 1 1
 ```
-
-Result (MI100 @ dynammic freq, 46TFlops peak FP32)
-```
-a_ms_ks: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
-b_ks_ns: dim 4, lengths {32, 64, 32, 64}, strides {128, 1, 524288, 4096}
-c_ms_ns: dim 4, lengths {30, 128, 32, 64}, strides {524288, 4096, 128, 1}
-launch_and_time_kernel: grid_dim {240, 1, 1}, block_dim {256, 1, 1}
-Warm up 1 time
-Start running 10 times...
-Perf: 0.843286 ms, 38.1985 TFlops, 94.5014 GB/s, DeviceContractionMultipleD_Xdl_CShuffle<256, 256, 128, 16, 4, 4>
-```
--- a/example/30_grouped_conv_fwd_multiple_d/README.md
+++ b/example/30_grouped_conv_fwd_multiple_d/README.md
--- a/example/34_batchnorm/batchnorm_infer_impl.hpp
+++ b/example/34_batchnorm/batchnorm_infer_impl.hpp
--- a/example/44_elementwise_permute/CMakeLists.txt
+++ b/example/44_elementwise_permute/CMakeLists.txt
--- a/example/44_elementwise_permute/elementwise_permute.cpp
+++ b/example/44_elementwise_permute/elementwise_permute.cpp
--- a/example/44_elementwise_permute/elementwise_permute_3d.cpp
+++ b/example/44_elementwise_permute/elementwise_permute_3d.cpp
--- a/example/44_elementwise_permute/elementwise_permute_4D_fp16_2d.cpp
+++ b/example/44_elementwise_permute/elementwise_permute_4D_fp16_2d.cpp
--- a/example/46_gemm_add_multiply/README.md
+++ b/example/46_gemm_add_multiply/README.md
--- a/example/59_grouped_gemm_multi_ABD/CMakeLists.txt
+++ b/example/59_grouped_gemm_multi_ABD/CMakeLists.txt
--- a/example/59_grouped_gemm_multi_ABD/grouped_gemm_multi_abd_xdl_fixed_nk_bias_bf16_i8.cpp
+++ b/example/59_grouped_gemm_multi_ABD/grouped_gemm_multi_abd_xdl_fixed_nk_bias_bf16_i8.cpp
--- a/example/59_grouped_gemm_multi_ABD/grouped_gemm_multi_abd_xdl_fixed_nk_bias_fp16.cpp
+++ b/example/59_grouped_gemm_multi_ABD/grouped_gemm_multi_abd_xdl_fixed_nk_bias_fp16.cpp
--- a/example/60_gemm_multi_ABD/CMakeLists.txt
+++ b/example/60_gemm_multi_ABD/CMakeLists.txt
 add_example_executable(example_gemm_multi_ABD_xdl_fp16 gemm_multi_ABD_xdl_fp16.cpp)
+add_example_executable(example_gemm_multi_ABD_xdl_bias_fastgelu_bf16_i8 gemm_multi_ABD_xdl_bias_fastgelu_bf16_i8.cpp)
+add_example_executable(example_gemm_multi_ABD_xdl_multiply_bias_fastgelu_bf16_i8 gemm_multi_ABD_xdl_multiply_bias_fastgelu_bf16_i8.cpp)
+add_example_executable(example_gemm_multi_ABD_xdl_fastgelu_bf16_i8 gemm_multi_ABD_xdl_fastgelu_bf16_i8.cpp)
--- a/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_bias_fastgelu_bf16_i8.cpp
+++ b/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_bias_fastgelu_bf16_i8.cpp
--- a/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_fastgelu_bf16_i8.cpp
+++ b/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_fastgelu_bf16_i8.cpp
--- a/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_fp16.cpp
+++ b/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_fp16.cpp
--- a/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_multiply_bias_fastgelu_bf16_i8.cpp
+++ b/example/60_gemm_multi_ABD/gemm_multi_ABD_xdl_multiply_bias_fastgelu_bf16_i8.cpp
--- a/example/61_contraction_multi_ABD/contraction_multi_ABD_xdl_fp16.cpp
+++ b/example/61_contraction_multi_ABD/contraction_multi_ABD_xdl_fp16.cpp
--- a/example/62_convnd_activ/binary/CMakeLists.txt
+++ b/example/62_convnd_activ/binary/CMakeLists.txt
--- a/example/62_convnd_activ/binary/convnd_bwd_weight_xdl_bilinear_residual_fp16.cpp
+++ b/example/62_convnd_activ/binary/convnd_bwd_weight_xdl_bilinear_residual_fp16.cpp