Simple direct solver for vortex particle methods using HIP for GPU

Cross platform, cross vendor GPU code without the hassle of external tools or compilers.

Wrapper tool to convert CUDA source code to HIP code and compile it with HIPCC. Useful for learning CUDA programming using AMD devices..

Simple graphing calculator using HIP

Example cross-platform C++ libraries that can be either build with HIP or CUDA runtimes

Simple C++ program to calculate the distance between a vector and each vector within a set using AMD HIP

Speculative task-based runtime system

GPU Implementation of Newton Fractal Generator with Benchmarking

Stencil code for AMD GPUs

KMM: parallel dataflow scheduler and efficient memory management for multi-GPU platforms

the small raytracer project based on "Ray Tracing in One Weekend" using HIP

C/C++ wrapper library for fft-related computations on various targets

CPU and GPU (using HIP) implementations of phase pattern generators for use with spatial light modulators

Write GPU program with RAII

SGEMM implementations in HIP for NVIDIA / AMD GPUs

test code

The Task-Aware HIP (TAHIP) provides interoperability support between task-based programming models and HIP which enables the taskification of HIP operations and kernels on accelerators

A simple ONNX engine with portability in mind

YOLOv8 using onnx.cpp, without ONNX Runtime or OpenCV

Gromacs that can be accelerated at AMD GPU