A compiler for NET CIL for GP-GPU computing in .NET languages.
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
Campy.Compiler
Campy.Find
Campy.Graphs
Campy.Meta
Campy.Peek
Campy.Runtime
Campy.Utils
Campy
ConsoleApp1
ConsoleApp4
Tests
nuget
.gitignore
Campy.sln
Campy.sln.DotSettings.user
README.md
Runtime.sln
clean-completely.sh
demo.sh
kens-campy-link-9-1.sh
kens-campy-nvcc-9-1.sh
license

README.md

Campy

Welcome! Campy is a compiler, framework, and API for GP-GPU computing in .NET languages. With Campy, one writes GP-GPU code without the usual boilerplate code you see in CUDA/OpenCL, but instead use a simplified model of a multiprocessor GPU/CPU computer, freeing the developer to focus exclusively on the algorithm. The API compiles and runs CIL code into native GPU code using LLVM. Supported are value types, reference types, methods, generics, lambdas, delegates, and closures. Other C#/GPU projects exist, but do not offer a clean, boilerplate-free interface, nor support C# beyond value types. Campy provides a base class library (BCL) for the GPU, implementing strings, arrays, generics, exceptions, delegates, dynamically memory allocation, reflection, and native calls.

This project is in the early stage of development. Releases are currently for demonstration purposes. The only available method is essentially Campy.Parallel.For(), and it will undergo wild changes in signature as I refine the programming model. There are also several debugging output switches (see example below).

public static void Campy.Parallel.For(int number_of_threads, _Kernel_type kernel)
public delegate void Campy.Types._Kernel_type(Index idx);

Targets

  • Windows 10 (x64), Ubuntu 16.04 (x64), Net Framework >= 4.6.1, Net Core 2.0; CUDA GPU Toolkit 9.2.148; Maxwell or better GPU.

Campy under a minute

(Make sure to install Net Core 2.0, https://www.microsoft.com/net/learn/get-started/windows.)

One step install: copy and paste the following code in a Bash shell.


#!/bin/bash
mkdir test
cd test
dotnet new console
cat - << HERE > Program.cs
namespace test
{
    class Program
    {
        static void Main(string[] args)
        {
            int n = 4;
            int[] x = new int[n];
            Campy.Parallel.For(n, i => x[i] = i);
            for (int i = 0; i < n; ++i)
                System.Console.WriteLine(x[i]);
        }
    }
}
HERE
dotnet add package Campy
dotnet build
unameOut="$(uname -s)"
case "${unameOut}" in
    Linux*)
	dotnet publish -r ubuntu.16.04-x64
	cd bin/Debug/netcoreapp2.1/ubuntu.16.04-x64/publish/
        ./test
	;;
    Darwin*)
	echo Cannot target Mac yet.
	exit 1
	;;
    CYGWIN*)
	dotnet publish -r win-x64
	cd bin/Debug/netcoreapp2.1/win-x64/publish/
        ./test.exe
	;;
    MINGW*)
	dotnet publish -r win-x64
	cd bin/Debug/netcoreapp2.1/win-x64/publish/
        ./test.exe
	;;
    *)
	echo Unknown machine.
	exit 1
	;;
esac
echo Output should be four lines of integers, 0 to 3.
Once an app is "published" as a self-contained deployment, it is completely sufficient.

Additional examples in Campy test area (https://github.com/kaby76/Campy/tree/master/Tests), including Reduction, various sorting algorithms, FFT, etc.

Notes on Building Campy from Scratch

  • Make sure to install VS 2017 15.8.4.
  • Net SDK 2.1.402
  • Make sure to install NVIDIA GPU Toolkit 9.2.148.
  • Through VS 2017, build Runtime.sln DEBUG first. It will prompt to upgrade--do not! Click on cancel and continue.
  • Through VS 2017, build Campy.sln DEBUG.