Compiling AscendNPU IR
The bishengir-compile tool can compile AscendNPU IR into a binary file that can be executed on the Ascend AI Processor.
Compilation Options
Common compilation options are described as follows.
Option |
Required (Yes/No) |
Description |
|---|---|---|
--help |
No |
Displays the help information. |
-o <file> |
No |
Specifies the name and location of the output file. |
--enable-triton-kernel-compile |
No |
Enables Triton-related compilation optimization. The input IR must be the MLIR converted by the triton-shared tool. |
--enable-hfusion-compile |
No |
Enables HFusion dialect-level compilation optimization. |
--enable-hivm-compile |
Yes |
Enables HIVM dialect-level compilation optimization. This option is enabled by default. |
--enable-auto-multi-buffer |
No |
Enables automatic ping pong buffer optimization of the Ascend AI Processor on-chip buffer. |
Compile Command
During compilation, the input.mlir file is compiled and optimized by the HFusion and HIVM dialects to generate the binary file outputFile.o.
The compile command is as follows:
1 | bishengir-compile --enable-hfusion-compile=true --enable-hivm-compile=true input.mlir -o outputFile |