Skip to content

Latest commit

 

History

History
53 lines (37 loc) · 1.9 KB

README.rst

File metadata and controls

53 lines (37 loc) · 1.9 KB

Loopy lets you easily generate the tedious, complicated code that is necessary to get good performance out of GPUs and multi-core CPUs.


Places on the web related to Loopy:


Loopy's core idea is that a computation should be described simply and then transformed into a version that gets high performance. This transformation takes place under user control, from within Python.

It can capture the following types of optimizations:

  • Vector and multi-core parallelism in the OpenCL/CUDA model
  • Data layout transformations (structure of arrays to array of structures)
  • Loopy Unrolling
  • Loop tiling with efficient handling of boundary cases
  • Prefetching/copy optimizations
  • Instruction level parallelism
  • and many more

Loopy targets array-type computations, such as the following:

  • dense linear algebra,
  • convolutions,
  • n-body interactions,
  • PDE solvers, such as finite element, finite difference, and Fast-Multipole-type computations

It is not (and does not want to be) a general-purpose programming language.

Loopy is licensed under the liberal MIT license and free for commercial, academic, and private use. All of Loopy's dependencies can be automatically installed from the package index after using:

pip install loo.py

In addition, Loopy is compatible with and enhances pyopencl.