rocFFT

rocFFT copied to clipboard

resolves #___ Summary of proposed changes: - Cherry Pick Fix for old version rocm include/lib folders not removed on upgrade commit - Remove include and lib directory from CPACK_RPM_EXCLUDE_FROM_AUTO_FIL… -

arvindcheru

RFC: RTC kernel cache file behaviour

1

This is a request for comment (RFC) regarding how rocFFT organizes the kernel cache file. rocFFT is starting to make move use of runtime compilation (RTC) in order to allow...

malcolmroberts

rtc: use gridDim.x for everything in real-complex kernels

1

Fixes launch errors on very large odd-length real-complex transforms, since gridDim.y and gridDim.z are limited to 64k.

af-ayala

Allow multi-gpu testing on systems without direct peer access

1

* Allow multi-gpu testing on systems without direct peer access * remove CheckAccess function

af-ayala

Performance ISSUE: Slow performance of rocfft comaped to cufft in MI200 series accelerators vs A100 GPUs

1

### Actual Issue : I was testing the performance of rocFFT on MI 200 series accelerators GPU vs cuFFT performance on A100 GPUs for single GPU only. ### What is...

manver-iitk

Re: https://github.com/ROCm/rocFFT/blob/1680fbde578bbc65d1abfbaadf53e49dd5e0494c/library/src/rtc_compile.cpp This is what CPU utilization looks like:  Somehow the usage is artificially capped at 50/70% of each core rather than just getting the job done.

alerque

Add CMake for documentation

6

Summary of proposed changes: Add CMake support for documentation.

matyas-streamhpc

Require rocFFT use hip-config.cmake to find HIP

1

Sometimes rocFFT use FindHIP.cmake which is not desired -- it does not have hip::device and hip::host target imported and result in cmake generating error. Reference: https://bugs.gentoo.org/932155

littlewu2508

malcolmroberts

build(deps): bump rocm-docs-core from 1.6.2 to 1.7.2 in /docs/sphinx

2

Bumps [rocm-docs-core](https://github.com/ROCm/rocm-docs-core) from 1.6.2 to 1.7.2. Release notes Sourced from rocm-docs-core's releases. v1.7.2 (2024-08-21) Fix projects.py: Update pattern matching for old release announcement [main e606899] bump: version 1.7.1 → 1.7.2...

dependabot[bot]