accelerate: An embedded language for accelerated array processing

Data.Array.Accelerate defines an embedded array language for computations for high-performance computing in Haskell. Computations on multi-dimensional, regular arrays are expressed in the form of parameterised collective operations, such as maps, reductions, and permutations. These computations may then be online compiled and executed on a range of architectures.

A simple example

As a simple example, consider the computation of a dot product of two vectors of floating point numbers:

dotp :: Acc (Vector Float) -> Acc (Vector Float) -> Acc (Scalar Float)
dotp xs ys = fold (+) 0 (zipWith (*) xs ys)

Except for the type, this code is almost the same as the corresponding Haskell code on lists of floats. The types indicate that the computation may be online-compiled for performance – for example, using Data.Array.Accelerate.CUDA it may be on-the-fly off-loaded to the GPU.

Available backends

Currently, there are two backends:

1. An interpreter that serves as a reference implementation of the intended semantics of the language, which is included in this package.

2. A CUDA backend generating code for CUDA-capable NVIDIA GPUs: http://hackage.haskell.org/package/accelerate-cuda

Several experimental and/or incomplete backends also exist. If you are interested in helping finish these, please contact us.

1. Cilk/ICC and OpenCL: https://github.com/AccelerateHS/accelerate-backend-kit

2. Another OpenCL backend: https://github.com/HIPERFIT/accelerate-opencl

3. A backend to the Repa array library: https://github.com/blambo/accelerate-repa

Additional components

The following support packages are available:

1. accelerate-cuda: A high-performance parallel backend targeting CUDA-enabled NVIDIA GPUs. Requires the NVIDIA CUDA SDK and, for full functionality, hardware with compute capability 1.2 or greater. See the table on Wikipedia for supported GPUs: http://en.wikipedia.org/wiki/CUDA#Supported_GPUs

2. accelerate-examples: Computational kernels and applications showcasing Accelerate, as well as performance and regression tests.

3. accelerate-io: Fast conversion between Accelerate arrays and other formats, including Repa arrays.

4. accelerate-fft: Computation of Discrete Fourier Transforms.

Install them from Hackage with cabal install PACKAGE

Examples and documentation

Haddock documentation is included in the package, and a tutorial is available on the GitHub wiki: https://github.com/AccelerateHS/accelerate/wiki

The accelerate-examples package demonstrates a range of computational kernels and several complete applications, including:

• An implementation of the Canny edge detection algorithm

• An interactive Mandelbrot set generator

• A particle-based simulation of stable fluid flows

• An n-body simulation of gravitational attraction between solid particles

• A cellular automata simulation

• A "password recovery" tool, for dictionary lookup of MD5 hashes

Mailing list and contacts

• Mailing list: accelerate-haskell@googlegroups.com (discussion of both use and development welcome).

• Sign up for the mailing list here: http://groups.google.com/group/accelerate-haskell

• Bug reports and issue tracking: https://github.com/AccelerateHS/accelerate/issues

Release notes

• 0.13.0.0: New array fusion optimisation. New foreign function interface for array and scalar expressions. Additional Prelude-like functions. New example programs. Bug fixes and performance improvements.

• 0.12.0.0: Full sharing recovery in scalar expressions and array computations. Two new example applications in package accelerate-examples: Real-time Canny edge detection and fluid flow simulator (both including a graphical frontend). Bug fixes.

• 0.11.0.0: New Prelude-like functions zip*, unzip*, fill, enumFrom*, tail, init, drop, take, slit, gather*, scatter*, and shapeSize. New simplified AST (in package accelerate-backend-kit) for backend writers who want to avoid the complexities of the type-safe AST.

• 0.10.0.0: Complete sharing recovery for scalar expressions (but currently disabled by default). Also bug fixes in array sharing recovery and a few new convenience functions.

• 0.9.0.0: Streaming, precompilation, Repa-style indices, stencils, more scans, rank-polymorphic fold, generate, block I/O & many bug fixes.

• 0.8.1.0: Bug fixes and some performance tweaks.

• 0.8.0.0: replicate, slice and foldSeg supported in the CUDA backend; frontend and interpreter support for stencil. Bug fixes.

• 0.7.1.0: The CUDA backend and a number of scalar functions.

Modules

[Index]

Flags

Automatic Flags

Name	Description	Default
debug	Enable tracing message flags. These are read from the command-line arguments, which is convenient but may cause problems interacting with the user program, so are disabled by default. The available options are: -ddump-sharing: print sharing recovery information -ddump-simpl-stats: dump statistics counts from the simplifier phase -ddump-simpl-iterations: dump the program after each iteration of the simplifier -dverbose: other, uncategorised messages	Enabled
more-pp	Enable HTML and Graphviz pretty printing.	Disabled
bounds-checks	Enable bounds checking	Enabled
unsafe-checks	Enable bounds checking in unsafe operations	Disabled
internal-checks	Enable internal consistency checks	Disabled

Use -f <flag> to enable a flag, or -f -<flag> to disable that flag. More info

Downloads

• accelerate-0.13.0.0.tar.gz [browse] (Cabal source package)
• Package description (as included in the package)

Maintainer's Corner

Package maintainers

• ManuelChakravarty, TrevorMcDonell

For package maintainers and hackage trustees

• edit package information

Candidates

• 1.3.0.0