stdgpu
A C++17 library providing efficient STL-like data structures (vector, unordered_map, etc.) for GPU programming with CUDA, OpenMP, and HIP backends.
What is stdgpu?
stdgpu is a C++ library that provides efficient STL-like data structures for GPU programming, enabling developers to manage data on GPUs with familiar container interfaces. It solves the problem of flexible and reliable data management in GPU algorithms by offering generic containers like vector, unordered_map, and unordered_set that work directly in GPU memory. The library supports multiple backends including CUDA, OpenMP, and HIP, making it versatile for different parallel computing environments.
C++ developers and researchers working on GPU-accelerated applications, particularly in fields like real-time 3D reconstruction, computer vision, scientific computing, and high-performance computing who need efficient data structures on the GPU.
Developers choose stdgpu because it provides a specialized, lightweight container library for GPUs that complements existing algorithm-focused libraries like thrust. Its unique selling point is the dual interface design allowing both high-level agnostic code and low-level native kernel integration, along with multi-backend support for maximum flexibility in GPU programming.
Overview
stdgpu: Efficient STL-like Data Structures on the GPU
Use Cases
Best For
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
