docs related add NonuniformFFTs.jl

flatironinstitute · Dec 9, 2024 · 4565d4b · 4565d4b
1 parent 244dd18
commit 4565d4b
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 5 deletions.
diff --git a/docs/julia.rst b/docs/julia.rst
@@ -1,11 +1,13 @@
+.. _julia:
+
 Julia interfaces (CPU and GPU)
 ==============================
 
 Principal author Ludvig af Klinteberg and others have built and maintain `FINUFFT.jl <https://github.com/ludvigak/FINUFFT.jl>`_, an interface from the `Julia <https://julialang.org/>`_ language. This official Julia package supports 32-bit and 64-bit precision, now on both CPU and GPU (via `CUDA.jl`), via a common interface.
 The Julia package installation automatically downloads pre-built CPU binaries of the FINUFFT library for Linux, macOS, Windows and FreeBSD (for a full list see `finufft_jll <https://github.com/JuliaBinaryWrappers/finufft_jll.jl>`_), and the GPU binary for Linux (see `cufinufft_jll <https://github.com/JuliaBinaryWrappers/cufinufft_jll.jl>`_).
 
 `FINUFFT.jl` has itself been wrapped as part of `NFFT.jl <https://juliamath.github.io/NFFT.jl/dev/performance/>`_, which contains an "abstract" interface
-to any NUFFT in Julia, with FINUFFT as an example.
+to any NUFFT in Julia, with FINUFFT as an example. This was by Tobias Knopp and coworkers, starting around 2022.
 Their
 `performance comparison page <https://juliamath.github.io/NFFT.jl/dev/performance/>`_
 show that FINUFFT matches their native Julia implementation for speed of type 1

diff --git a/docs/related.rst b/docs/related.rst
@@ -16,7 +16,9 @@ Other recommended NUFFT libraries
 
 - `PyNUFFT <https://github.com/jyhmiinlin/pynufft>`_ Python code supporting CPU and GPU operation. We have not compared against FINUFFT yet.
 
-
-Also see the summary of library performances in our paper [FIN] in the
-:ref:`references <refs>`.
-
+- `NonuniformFFTs.jl <https://jipolanco.github.io/NonuniformFFTs.jl/dev/>`_ native Julia code for types 1 and 2 only (CPU and GPU via KernelAbstractions), by Juan Polanco, 2024. Close to our CPU performance, and can beat it in the case of real data via a custom real transform. On the GPU claims their shared-memory type 1 implementation beats ours; to be investigated further. Has a good `benchmarks page <https://jipolanco.github.io/NonuniformFFTs.jl/dev/benchmarks/>`_ comparing (cu)FINUFFT at 6-digit accuracy, CPU and GPU.
+
+- `NFFT.jl <github.com/JuliaMath/NFFT.jl>`_ native Julia implementation for type 1 and 2 only, by Tobias Knopp and coworkers, starting around 2022. See :ref:`page on Julia <julia>`.
+
+
+A comparison of some library performances (as of 2019) was in our paper [FIN] in the :ref:`references <refs>`.