CUDAfy.NET 1.25.4963.10126

CUDAfy is an SDK that permits general purpose programming of NVIDIA CUDA Graphics Processing Units (GPUs) and OpenCL devices completely from within the Microsoft .NET framework.  Device code and host code can be written in .NET using standard .NET types. There are also wrappers for the CUDA Random, FFT and BLAS libraries.

For CUDA development the NVIDIA CUDA 5.5 Toolkit must be installed.  For OpenCL the relevant SDK from Intel, AMD, NVIDIA or Altera must be installed.

For first time users it is recommended to download the installer / set-up tool from

There is a newer version of this package available.
See the version list below for details.
Install-Package CUDAfy.NET -Version 1.25.4963.10126
dotnet add package CUDAfy.NET --version 1.25.4963.10126
<PackageReference Include="CUDAfy.NET" Version="1.25.4963.10126" />
For projects that support PackageReference, copy this XML node into the project file to reference the package.
paket add CUDAfy.NET --version 1.25.4963.10126
The NuGet Team does not provide support for this client. Please contact its maintainers for support.
#r "nuget: CUDAfy.NET, 1.25.4963.10126"
#r directive can be used in F# Interactive, C# scripting and .NET Interactive. Copy this into the interactive tool or source code of the script to reference the package.
// Install CUDAfy.NET as a Cake Addin
#addin nuget:?package=CUDAfy.NET&version=1.25.4963.10126

// Install CUDAfy.NET as a Cake Tool
#tool nuget:?package=CUDAfy.NET&version=1.25.4963.10126
The NuGet Team does not provide support for this client. Please contact its maintainers for support.

Release Notes

Fix: Typo in xml comment of GThread.
Fix: Use of eCudafyAddressSpace attribute on parameters (needed for OpenCL).
Fix: Shared memory arrays of structs not translated correct for OpenCL.
Fix: Cudafying reported by some users to appear to freeze when calling nvcc.
Chg: Make HostAllocate for OpenCL 4096 byte aligned.
Chg: OpenCL Linux so name
Add: LinuxDebug configuration to Cloo.VS2010.csproj
Add: More flexible Cudafy method using CompileProperties (testing to be   finalized).
Add: Dynamic Parallelism support (alpha).
Add: Ability to GetDevice based on architecture
Add: Binary (cubin) support.
Chg: Cudafy.Host.Unit tests.
Chg: Update cudafybyexample
Chg: If no arch specified fall back on CudafyModes.Lanaguage prop
Add: CUDA 5.5 (required for RT API)
Add: New *.cdfy files (not backward compatible)
Add: Clone method to CudafyModule to allow easier loading of same module to   multiple GPUs.
Add: Support GetDeviceProperties advanced for both current and previous CUDA   release.
Add: .NET 3.5 support.
Fix: Delete generated code.
Add: CUDA 5.5 support.


This package has no dependencies.

NuGet packages

This package is not used by any NuGet packages.

GitHub repositories

This package is not used by any popular GitHub repositories.

Version History

Version Downloads Last updated
1.29.5576.13786 9,675 4/8/2015
1.27.5314.36419 1,563 8/27/2014
1.25.4963.10126 3,889 8/21/2013
1.22.4875.21265 1,033 5/8/2013
1.12.4695.21111 1,075 12/1/2012
1.9.4495.35683 1,208 5/9/2012
1.8.4427.36820 1,072 2/14/2012
1.7.4364.27657 1,051 12/22/2011
1.5.4315.23069 1,130 10/26/2011
1.4.4209.28745 1,257 7/12/2011
1.3.4188.30556 1,196 6/21/2011