rocFFT-1.0.10 for ROCm 4.1.0

Added

• Explicitly specify MAX_THREADS_PER_BLOCK through _launch_bounds for all kernels.
• Switch to new syntax for specifying AMD GPU architecture names and features.

Optimizations

• Optimized C2C/R2C 3D 64, 81, 100, 128, 200, 256 cube sizes.
• Improved performance of the standalone out-of-place transpose kernel.
• Optimized 1D length 40000 C2C case.
• Enabled radix-7 for size 336.
• New radix-11 and radix-13 kernels; used in length 11 and 13 (and some of their multiples) transforms.

Changed

• rocFFT now automatically allocates a work buffer if the plan requires one but none is provided.
• An explicit rocfft_status_invalid_work_buffer error is now returned when a work buffer of insufficient size is provided.
• Updated online documentation.
• Updated debian package name version with separated '_'.
• Adjusted accuracy test tolerances and how they are compared.

Fixed

• Fixed 4x4x8192 accuracy failure.

Known Issues

• None

rocFFT-1.0.8 for ROCm 4.0.0

New Features

• No new features

Known Issues

• None

rocFFT-1.0.8 for ROCm 3.10.0

New Features

• Added deprecation warning for hipfft.h
• Optimized case 1D 10000 C2C
• Fixed SBCC/SBRC non-unit stride batch issue
• Updated README and added BUILD_CLIENTS_ALL
• Improved test infrastructure

Known Issues

• None

rocFFT-1.0.7 for ROCm 3.9.0

New Features

• Optimized 3D C2C cases
• Optimized large pow-of-2 cases
• Fused R2C/C2R even length post/pre-process with transpose kernel
• Enabled 2D single kernel for pow3 and mixed pow2/pow3
• Added radix-7 optimazation
• Added 1D batch-paired R2C transform
• Added -mno-xnack -mno-sram-ecc build flags
• Fixed repo singleton destrory issue
• Fixed 1D non-unit stride issue

Known Issues

• None

rocFFT-1.0.6 for ROCm 3.8.0

New Features

• optimized 1D C2C sizes: 6561, 10000
• optimized 2D pow-of-2 small cases
• optimized 2D rectangular cases: 256x264, 256x272, … 256x320, 256x4096, 256x8192
• fixed buffer assigment bug of large 1D cases* improved test and sample code

Known Issues

• None

rocFFT-1.0.5 for ROCm 3.7.0

New Features

• Optimized 2D C2C middle sizes with fused 2 kernels for pow-of-2
• Change package dependency to hip-rocclr
• Fixed build issue with C++ 17
• Improved test infrastructure

Known Issues

• None

rocFFT-1.0.4 for ROCm 3.6.0

New Features

• Fixed non-unit stride issue of 1D middle size
• Updated client package installation path
• Improved internal device memory usage check
• Improved log
• Improved test infrastructure

Known Issues

• None

rocFFT-1.0.3 for ROCm 3.5.0

New Features

• Switched to hip-clang as default compliler and deprecated hcc build
• Improved hip-nvcc build
• Support static lib build
• Improved tests

Known Issues

None

rocFFT-1.0.3 for ROCm 3.5.0

New Features

• Switched to hip-clang as default compliler and deprecated hcc build
• Improved hip-nvcc build
• Support static lib build
• Improved tests

Known Issues

None

rocFFT-1.0.2 release for ROCM-3.3

• Supported planar(split) format
• Updated build instructions in README
• Removing ROCm libs from dependency installation and updates to RUNPATH
• Improved accuracy, rider, sample, CI tests, and added dyna-rider test