mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-01 12:11:19 +00:00
92c67a824f
* GH-2368 Adding a basic glossary GH-2368 Minor edits GH-2368 Adding missing READMEs and standardization. resolving readme updates GH-2368 Minor improvements to documentation. Improving some readmes. Further improvement for readmes. Cleaned up the documentation in 'client_example' (#2468) Update for PR Update ACRONYMS.md to remove trivial terms Update ACRONYMS.md to provide detailed explanations for BF16 and BF8 formats Apply suggestion from @spolifroni-amd Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com> Apply suggestion from @spolifroni-amd Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com> Update README.md to clarify CK Tile API description and remove outdated references to the Tile Engine. revise 37_transpose readme revise 36_copy readme Remove references to the Tile Engine in README files for 19_gemm_multi_d and 35_batched_transpose, and update distribution links for clarity. Remove references to the Tile Engine in multiple README files and update distribution links for consistency and clarity. Remove references to the Tile Engine in README files across multiple examples * GH-2368 Adding a basic glossary GH-2368 Minor edits GH-2368 Adding missing READMEs and standardization. resolving readme updates GH-2368 Minor improvements to documentation. Improving some readmes. Further improvement for readmes. Cleaned up the documentation in 'client_example' (#2468) Update for PR Update ACRONYMS.md to remove trivial terms Update ACRONYMS.md to provide detailed explanations for BF16 and BF8 formats Apply suggestion from @spolifroni-amd Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com> Apply suggestion from @spolifroni-amd Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com> Update README.md to clarify CK Tile API description and remove outdated references to the Tile Engine. revise 37_transpose readme revise 36_copy readme Remove references to the Tile Engine in README files for 19_gemm_multi_d and 35_batched_transpose, and update distribution links for clarity. Remove references to the Tile Engine in multiple README files and update distribution links for consistency and clarity. Remove references to the Tile Engine in README files across multiple examples Refine README files by removing outdated references to the Tile Engine * Updates based on PR feedback 1 * Updates based on PR feedback 2 * Updates based on PR feedback 3 * Updates based on PR feedback 4 * Updates based on PR feedback 5 * Updates based on PR feedback 6 * Updates based on PR feedback 7 * Updates based on PR feedback 8 * Content Modification of CK Tile Example * Modify the ck_tile gemm config --------- Co-authored-by: AviralGoelAMD <aviral.goel@amd.com> Co-authored-by: ThomasNing <thomas.ning@amd.com>
Add + RMSNorm2D + Rowwise Dynamic Quantization (RDQuant) with CK Tile
This example demonstrates a fused kernel for elementwise addition, 2D RMSNorm, and rowwise dynamic quantization using the CK Tile programming model. This pattern is common in LLMs for efficient normalization and quantized inference.
Algorithm and Math
Given input X and residual R:
- Elementwise Add:
Z = X + R - RMSNorm:
\text{rms}(Z) = \sqrt{\frac{1}{N} \sum_{i=1}^N Z_i^2 + \epsilon},Y_i = \frac{Z_i}{\text{rms}(Z)} \cdot \gamma_i - Rowwise Dynamic Quantization:
- For each row,
s = \max(|Y|) / 127 Q_i = \text{round}(Y_i / s),Q_i \in \text{int8}
- For each row,
Output:
- Quantized tensor
Q(int8) - Per-row scale
s(fp32)
Tile Programming Model
- Tiles: Each thread block processes a tile (row or block).
- Tile Engine: Loads tiles, performs add, RMSNorm, and quantization.
- Pipeline: Modular, can be extended for further fusion.
Build & Run
# in the root of ck_tile
mkdir build && cd build
sh ../script/cmake-ck-dev.sh ../ <arch> # you can replace this <arch> to gfx90a, gfx942...
make tile_add_rmsnorm2d_rdquant_fwd -j`nproc`
This will result in an executable build/bin/tile_add_rmsnorm2d_rdquant_fwd
Arguments
args:
-m m dimension (default:3328)
-n n dimension (default:4096)
-stride stride per row, if -1 then equal to n (default:-1)
-e epsilon (default:1e-5)
-save_x save rms(invrms) or not. set to 1 in training case (default:1)
-v cpu validation or not (default:1)
-kname print kernel name or not (default:1)
-prec precision (default:fp16)
-quant precision (default:int8)
-warmup cold iter (default:5)
-repeat hot iter (default:20)
-json 0: No Json, 1: Dump Results in Json format (default:0)
-jsonfile json file name to dump results (default:add_rmsnorm2d_rdquant_fwd.json)
Source Structure
- Kernel:
add_rmsnorm2d_rdquant_fwd.hpp(tile-programming kernel template) - Executable:
add_rmsnorm2d_rdquant_fwd.cpp,example_add_rmsnorm2d_rdquant_fwd.cpp - Build:
CMakeLists.txt,instances/,script/
Related CK Tile Examples
- 10_rmsnorm2d: RMSNorm2D with tiles
- 12_smoothquant: SmoothQuant with tiles
- 02_layernorm2d: LayerNorm2D with tiles
For distribution, see include/ck_tile/tile_program/tile_distribution/.