mlir-dace
所属分类:人工智能/神经网络/深度学习
开发工具:C++
文件大小:676KB
下载次数:0
上传日期:2023-04-24 20:36:52
上 传 者:
sh-1993
说明: mlir dace,以数据为中心的mlir方言
(mlir-dace,Data-Centric MLIR dialect)
文件列表:
.clang-format (56, 2023-04-13)
CMakeLists.txt (2750, 2023-04-13)
LICENSE (1512, 2023-04-13)
design (0, 2023-04-13)
design\commands.tex (783, 2023-04-13)
design\file_compatibility.tex (360, 2023-04-13)
design\img (0, 2023-04-13)
design\img\consume.png (28982, 2023-04-13)
design\img\lib.png (31962, 2023-04-13)
design\img\map.png (41852, 2023-04-13)
design\img\multistate.png (17192, 2023-04-13)
** (42649, 2023-04-13)
design\img\state.png (13074, 2023-04-13)
design\img\state_in_out.png (14463, 2023-04-13)
design\img\symbol.png (30703, 2023-04-13)
design\img\tasklet.png (30034, 2023-04-13)
design\img\transient_array.png (24737, 2023-04-13)
design\main.pdf (266130, 2023-04-13)
design\main.tex (567, 2023-04-13)
design\mlir (0, 2023-04-13)
design\mlir\consume.mlir (516, 2023-04-13)
design\mlir\file_comp.mlir (262, 2023-04-13)
design\mlir\lib.mlir (219, 2023-04-13)
design\mlir\map.mlir (181, 2023-04-13)
design\mlir\multistate.mlir (219, 2023-04-13)
design\mlir\nested.mlir (308, 2023-04-13)
** (333, 2023-04-13)
design\mlir\state.mlir (39, 2023-04-13)
design\mlir\state_in_out.mlir (105, 2023-04-13)
design\mlir\stream.mlir (259, 2023-04-13)
design\mlir\symbol.mlir (155, 2023-04-13)
... ...
[![Tests](https://github.com/spcl/mlir-dace/actions/workflows/lit-test.yml/badge.svg)](https://github.com/spcl/mlir-dace/actions/workflows/lit-test.yml)
# MLIR-DaCe
MLIR-DaCe is a project aiming to bridge the gap between control-centric and data-centric intermediate representations.
By bridging these two groups of IRs, it allows the combination of control-centric and data-centric optimizations in optimization pipelines.
In order to achieve this, MLIR-DaCe provides a data-centric dialect in MLIR to connect the MLIR and DaCe frameworks.
## Building
This setup assumes that you have built LLVM and MLIR in `$BUILD_DIR` and installed them to `$PREFIX`. To build and launch the tests, run
```sh
mkdir build && cd build
cmake -G Ninja .. -DMLIR_DIR=$PREFIX/lib/cmake/mlir -DLLVM_EXTERNAL_LIT=$BUILD_DIR/bin/llvm-lit
cmake --build . --target check-s***-opt
```
To build the documentation from the TableGen description of the dialect operations, run
```sh
cmake --build . --target mlir-doc
```
**Note**: Make sure to pass `-DLLVM_INSTALL_UTILS=ON` when building LLVM with CMake in order to install `FileCheck` to the chosen installation prefix.
## Publication
The paper on DCIR can be found [here](https://dl.acm.org/doi/10.1145/3579990.3580018).
If you use MLIR-DaCe, cite us:
```bibtex
@inproceedings{mlir-dace,
author = {Ben-Nun, Tal and Ates, Berke and Calotoiu, Alexandru and Hoefler, Torsten},
title = {Bridging Control-Centric and Data-Centric Optimization},
year = {2023},
booktitle = {Proceedings of the 21st ACM/IEEE International Symposium on Code Generation and Optimization},
series = {CGO 2023}
}
```
近期下载者:
相关文件:
收藏者: