Hybridizer Documentation

Hybridizer is a compiler that transforms your C# code into high-performance GPU (CUDA) and CPU SIMD (AVX/NEON) code. Write once in C#, run everywhere — on GPU and CPU — with near-native performance.

New to Hybridizer?

Start with the Tutorials — a step-by-step learning path that takes you from installation to advanced GPU patterns:

1. Setup & Installation — Get your environment ready in 15 minutes
2. Your First Kernel — Write, build, and run a GPU kernel
3. Understanding the Result — Debug, verify, and optimize transfers
4. From CPU to GPU — Transform existing C# code for GPU
5. Working with Images — 2D kernels and stencil operations
6. Shared Memory & Reduction — The key GPU cooperative pattern

Already Know the Basics?

• Concepts — Architecture, compilation pipeline, data marshalling
• Examples — 9 complete working samples from the GitHub repo
• How-To Recipes — Port code, optimize, manage memory
• Reference — All attributes, CLI options, API index