Skip to content

parallax-compiler/parallax-docs

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Parallax Documentation

Official documentation for the Parallax GPU offload compiler.

Quick Links

What is Parallax?

Parallax is an automatic GPU offload compiler for C++ parallel algorithms. It transforms standard C++20 code into GPU-accelerated code using Vulkan, without requiring any source code modifications.

Key Features

  • 100% ISO C++20 compliant - Pure standard C++
  • Zero source changes - Just use std::execution::par
  • Automatic GPU offload - Compiler handles everything
  • Portable - Works on NVIDIA, AMD, Intel GPUs
  • High performance - 700+ million elements/second

Example

#include <vector>
#include <algorithm>
#include <execution>
#include <parallax/allocator.hpp>

int main() {
    std::vector<float, parallax::allocator<float>> data(1000000);

    // Runs on GPU automatically!
    std::for_each(std::execution::par, data.begin(), data.end(),
                 [](float& x) { x = x * 2.0f; });
}

No CUDA. No OpenCL. Just standard C++.

Documentation Structure

Getting Started

Guides

Reference

Advanced

Performance

Latest benchmarks on NVIDIA GTX 980M:

Dataset Throughput
1M elements 744 M elem/s
100K elements 228 M elem/s
10K elements 36.7 M elem/s

See Performance Guide for details.

Supported Algorithms

Algorithm Status Performance
std::for_each ✅ Production 744 M/s
std::transform ✅ Production 732 M/s
std::reduce ⏳ Planned -

System Requirements

  • GPU: Vulkan 1.2+ capable
  • Compiler: Clang 15+
  • OS: Linux, macOS, Windows
  • RAM: 4GB+ recommended

GitHub Pages

View the documentation online: https://YOUR_USERNAME.github.io/parallax-base

Building Documentation

cd parallax-docs
# View locally
python3 -m http.server 8000
# Visit http://localhost:8000

Contributing

See CONTRIBUTING.md for guidelines.

License

MIT License - see LICENSE

About

Documentation and website for Parallax

Resources

Code of conduct

Contributing

Stars

Watchers

Forks

Releases

No releases published

Sponsor this project

Packages

No packages published