rocFFT 1.0.15 for ROCm 5.0.1

rocFFT code for ROCm 5.0.1 is unchanged from rocFFT for ROCm 5.0.0. The library was rebuilt for the updated ROCm 5.0.1 stack.

rocFFT 1.0.15 for ROCm 5.0.0

Changed

• Re-aligned split device library into 4 roughly equal libraries.
• Implemented the FuseShim framework to replace the original OptimizePlan
• Implemented the generic buffer-assignment framework. The buffer assignment
  is no longer performed by each node. We designed a generic algorithm to
  test and pick the best assignment path.
  With the help of FuseShim, we can achieve more kernel-fusions as possible.
• Do not read the imaginary part of the DC and Nyquist modes for even-length
  complex-to-real transforms.

Optimizations

• Optimized twiddle-conjugation; complex-to-complex inverse transforms should have similar performance to foward transforms now.
• Improved performance of single-kernel small 2D transforms.

rocFFT 1.0.14 for ROCm 4.5.2

rocFFT code for ROCm 4.5.2 is unchanged from rocFFT for ROCm 4.5.0. The library was rebuilt for the updated ROCm 4.5.2 stack.

rocFFT 1.0.14 for ROCm 4.5.0

Changed

• Packaging split into a runtime package called rocfft and a development package called rocfft-devel. The development package depends on runtime. The runtime package suggests the development package for all supported OSes except CentOS 7 to aid in the transition. The suggests feature in packaging is introduced as a deprecated feature and will be removed in a future rocm release.

Optimizations

• Optimized SBCC kernels of length 52, 60, 72, 80, 84, 96, 104, 108, 112, 160,
  168, 208, 216, 224, 240 with new kernel generator.
• Improved many plans by removing unnecessary transpose steps.
• Optimized scheme selection for 3D problems.
  • Imposed less restrictions on 3D_BLOCK_RC selection. More problems can use 3D_BLOCK_RC and
    have some performance gain.
  • Enabled 3D_RC. Some 3D problems with SBCC-supported z-dim can use less kernels and get benefit.
  • Force --length 336 336 56 (dp) use faster 3D_RC to avoid it from being skipped by conservative
    threshold test.
• Optimized some even-length R2C/C2R cases by doing more operations
  in-place and combining pre/post processing into Stockham kernels.
• Added radix-17.

Fixed

• Fixed a few validation failures of even-length R2C inplace. 2D, 3D cubics sizes such as
  100^2 (or ^3), 200^2 (or ^3), 256^2 (or ^3)...etc. We don't combine the three kernels
  (stockham-r2c-transpose). We only combine two kernels (r2c-transpose) instead.
• Improved large 1D transform decompositions.

Added

• Added support for Windows 10 as a build target.
• Added new kernel generator for select fused-2D transforms.

rocFFT 1.0.12 for ROCm 4.3.1

Updated

• documentation notes that callbacks are experimental

rocFFT 1.0.12 for ROCm 4.3.0

Changed

Re-split device code into single-precision, double-precision, and miscellaneous kernels.

Fixed

• Fixed potential crashes in double-precision planar->planar transpose.

Added

• Added new kernel generator for select lengths. New kernels have
  improved performance.
• Added public rocfft_execution_info_set_load_callback and
  rocfft_execution_info_set_store_callback API functions to allow
  executing extra logic when loading/storing data from/to global
  memory during a transform.

Removed

• Removed R2C pair schemes and kernels.

Optimizations

• Optimized 2D/3D R2C 100 and 1D Z2Z 2500.
• Reduced number of kernels for 2D/3D sizes where higher dimension is 64, 128, 256.

Fixed

• Fixed potential crashes in 3D transforms with unusual strides, for
  SBCC-optimized sizes.

rocFFT-1.0.11 for ROCm 4.2.0

Optimizations

• Improved performance for single precision kernels exercising all except radix-2/7 butterfly ops.
• Minor optimization for C2R 3D 100, 200 cube sizes.
• Optimized some C2C/R2C 3D 64, 81, 100, 128, 200, 256 rectangular sizes.
• When factoring, test to see if remaining length is explicitly supported.
• Explicitly add radix-7 lengths 14, 21, and 224 to list of supported lengths.
• Optimized R2C 2D/3D 128, 200, 256 cube sizes.

Fixed

• Fixed potential crashes in small 3D transforms with unusual strides. (#311)
• Fixed potential crashes when executing transforms on multiple devices. (#310)

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