Difference between revisions of "Porting/ncnn"

From RCS Wiki
Jump to navigation Jump to search
(Finished: load_param_mem pybind)
(→‎Finished: python: document CMAKE_TOOLCHAIN_FILE env var)
 
(6 intermediate revisions by the same user not shown)
Line 5: Line 5:
 
* [https://github.com/Tencent/ncnn/pull/4853 Add POWER8 VSX toolchains]
 
* [https://github.com/Tencent/ncnn/pull/4853 Add POWER8 VSX toolchains]
 
* [https://github.com/Tencent/ncnn/pull/4924 load_param_mem pybind]
 
* [https://github.com/Tencent/ncnn/pull/4924 load_param_mem pybind]
 +
* [https://github.com/Tencent/ncnn/issues/5063 test_squeezenet failed under Lubuntu 16.04 PowerPC 32-bit]
 +
* [https://github.com/Tencent/ncnn/pull/5121 support big endian platform, add powerpc ci]
 +
* [https://github.com/Tencent/ncnn/pull/5174 Update POWER Clang version docs]
 +
* [https://github.com/Tencent/ncnn/pull/5178 Update Vulkan dependency docs]
 +
* [https://github.com/Tencent/ncnn/pull/5228 Document libomp-dev dependency]
 +
* [https://github.com/Tencent/ncnn/pull/5229 python: document CMAKE_TOOLCHAIN_FILE env var]
  
 
= In progress =
 
= In progress =
  
 
* CI missing for POWER9/Clang
 
* CI missing for POWER9/Clang
* SSE4.1 missing for POWER9/Clang on Fedora 38
 
** Caused by [[Porting/LLVM|LLVM]] issue 64664
 
 
* Replace SSE with native VSX
 
* Replace SSE with native VSX
 +
 +
= VSX Targets =
 +
 +
When running Real-ESRGAN in ncnn on POWER9, most CPU time (over 81%) is spent inside [https://github.com/Tencent/ncnn/blob/575098640c254be2208095254f3e5de412751447/src/layer/x86/convolution_3x3_winograd.h#L613 <code>gemm_transB_packed_tile</code> in <code>convolution_3x3_winograd.h</code>], which uses SSE2. This may be a good target for rewriting in native VSX.
  
 
= See Also =
 
= See Also =

Latest revision as of 13:24, 19 December 2023