Skip to content
No description or website provided.
Branch: master
Clone or download
Latest commit 2b35544 Oct 8, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
docs add loopinfo and unroll macro Mar 13, 2019
examples Remove CUDAnative in global scope dependency Apr 9, 2019
src Add signature function Oct 2, 2019
test fix inferrability of cpu shmem arrays Jun 18, 2019
.gitignore add gitignore Feb 1, 2019
.gitlab-ci.yml test on v1.2 Aug 11, 2019
.travis.yml add travis to verify Pkg.test Aug 11, 2019 add Feb 1, 2019
Project.toml add a note about using Cthulhu Oct 8, 2019
REQUIRE update invariant correctly Mar 18, 2019
bors.toml borsify gitlab ci Feb 18, 2019


Support for writing loop-based code that executes both on CPU and GPU


GPUifyLoops is a registered package, and can be installed using the Julia package manager.

(v1.1) pkg> add GPUifyLoops

Note: The current version of this package requires Julia 1.1.


Debugging failures to transforma a function for the GPU requires the use of Cthulhu.jl.

using Cthulhu
using GPUifyLoops

# @launch CUDA() f(args...)
descend(GPUifyLoops.signature(f, args...)...)


In order to test this package locally you need to do:

julia --project=test/gpuenv
julia> ]
(gpuenv) pkg> resolve
(gpuenv) pkg> instantiate

This will resolve the GPU environment, please do not checking changes to test/gpuenv/. Then you can run the tests with julia --project=test/gpuenv test/runtests.jl


GPUifyLoops.jl is licensed under MIT license.

You can’t perform that action at this time.