mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-14 10:09:41 +00:00
6c7e2d3a7e0787651db033cb50a8b35b6c5233bd
* start
* read for gridwise gemm
* add MakeBGridDescriptor_K0_N0_N1_N2_N3_K1
* add thread copy desc and register buffer
* add K0PerBlock dim
* add read global data
* finish gridwise gemm
* finish blockwise gemm
* add print data
* add smallest config
* add compare code for gridwis gemm
* fix NXdlPerWave
* fix k0perthread and gridewis gemm main loop
* remove b matrix lds alloc
* fix name
* add test code
* create b_grid_desc_k0_k1_k2_n0_n1_n2_n3_k3 from parameter
* add double register
* modify b_thread_desc_
* add float
* fp16 tag
* add tail for pipeline
* finish main loop
* optimize main loop
* start clear gridwise gemm
* clear code
* clear redundant code
* change file name
* change file name
* fix bug after merge develop
* fix input parameters
* using MultiK0 control b load data loop
* fix some config
* 4 buffer
* fix bug
* one can use
* change read order
* change buffer array to tuple
* change to 8 buffer
* interleave buffer load
* change to 16
* read 8 buffer
* add data buffer to template
* fix after merge develop(head file)
* format
* change to 4 buffer
* remove unnecessary lambda fun
[ROCm/composable_kernel commit: 10b3278b05]
Docker script
docker run \
-it \
--privileged \
--group-add sudo \
-w /root/workspace \
-v ${PATH_TO_LOCAL_WORKSPACE}:/root/workspace \
rocm/tensorflow:rocm5.1-tf2.6-dev \
/bin/bash
Install newer version of rocm-cmake
https://github.com/RadeonOpenCompute/rocm-cmake
Build
mkdir build && cd build
# Need to specify target ID, example below is gfx908 and gfx90a
cmake \
-D BUILD_DEV=OFF \
-D CMAKE_BUILD_TYPE=Release \
-D CMAKE_CXX_FLAGS=" --offload-arch=gfx908 --offload-arch=gfx90a -O3" \
-D CMAKE_CXX_COMPILER=/opt/rocm/bin/hipcc \
-D CMAKE_PREFIX_PATH=/opt/rocm \
-D CMAKE_INSTALL_PREFIX=${PATH_TO_CK_INSTALL_DIRECTORY} \
..
Build and Run Examples
make -j examples
Instructions for running each individual examples are under example/
Tests
make -j examples tests
make test
Build ckProfiler
make -j ckProfiler
Instructions for running ckProfiler are under profiler/
Install CK
make install
Using CK as pre-built kernel library
Instructions for using CK as a pre-built kernel library are under client_example/
Caveat
Kernel Timing and Verification
CK's own kernel timer will warn up kernel once, and then run it multiple times to get average kernel time. For some kernels that use atomic add, this will cause output buffer to be accumulated multiple times, causing verfication failure. To work around it, do not use CK's own timer and do verification at the same time. CK's own timer and verification in each example and ckProfiler can be enabled or disabled from command line.
Languages
C++
93.1%
Python
4.5%
CMake
1.5%
Shell
0.5%
Pawn
0.2%