# m4rs-mt/ILGPU

ILGPU JIT Compiler for high-performance .Net GPU programs

Repository: https://github.com/m4rs-mt/ILGPU
Canonical: https://ross.abutalabs.com/products/ilgpu
Homepage: http://www.ilgpu.net
Language: C#
License: NOASSERTION
License Family: other
Topics: ilgpu, jit, gpu, compiler, nvidia, amd, intel, ptx, cuda, opencl, dotnet, msil, cil, gpgpu, gpgpu-computing, cpu, kernels, parallel
Last push: 2026-07-30T16:51:09+00:00

## Health v2 (maintenance only)
Score: 68/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 95, release rhythm 16, longevity 100
- inputs: {"age_days": 3524, "days_push": 34, "days_rel": 417, "gap_med": 103, "n_releases_24m": 2}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1748, forks 140 (observed 2026-08-28T04:05:30.929883+00:00)

## What it is
ILGPU is a just-in-time compiler for high-performance GPU programs (kernels) written in .NET languages like C# and F#, entirely in C# with no native dependencies. It compiles kernels to CUDA, OpenCL, and PTX targets and includes a multi-threaded CPU accelerator for debugging and emulation.

## Use cases
- write GPU kernels in C# without learning CUDA or OpenCL
- run parallel computations on NVIDIA, AMD, or Intel GPUs from .NET
- debug GPU kernels on the CPU with a .NET debugger
- accelerate math-heavy workloads like N-body simulation
- use standard parallel algorithms like sorting and prefix sums on any accelerator
- prototype GPU code portably without native dependencies

## When to choose
- you want GPU acceleration from C# or F# with type-safe, unannotated kernels
- you need portable GPU code across CUDA and OpenCL vendors
- you want to debug kernels on the CPU using your normal .NET tooling
- you prefer a managed library with no native dependencies over writing C++ CUDA

## When to avoid
- you need maximum raw performance from hand-tuned CUDA or low-level PTX control
- your project is not on the .NET platform
- you need mature ecosystem support like deep learning frameworks built on top
- you require vendor-specific CUDA features not exposed by the compiler

## Facets
- artifact type: library
- maturity: active
- function: compiler, gpu-computing, machine-learning
- domain: gpu-computing, developer-tools, performance
- platform: cross-platform, windows, dotnet
- tags: jit-compiler, cuda, opencl, ptx, gpgpu, c-sharp, kernels, cpu-accelerator, parallel-computing, linux, macos, nvidia, amd, intel

## Member repositories
- m4rs-mt/ILGPU (main) score 68

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:30.929883+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:28:52.661012+00:00, confidence not recorded.
  - readme: https://github.com/m4rs-mt/ILGPU (fetched 2026-08-28T04:05:30.929883+00:00, sha 305d37522d1a)
  - homepage: http://www.ilgpu.net (fetched 2026-08-29T11:06:51.723706+00:00, sha 761958aa3fc0)
  - site_page: https://ilgpu.net/docs (fetched 2026-08-29T11:06:51.737156+00:00, sha 78d54a663c1b)
  - site_page: https://ilgpu.net/releases (fetched 2026-08-29T11:06:51.733011+00:00, sha 6e73a9af66e3)
- Data as of 2026-08-30T08:39:29.467469+00:00.