CVCUDA/CV-CUDA

CV-CUDA™ is an open-source, GPU accelerated library for cloud-scale image processing and computer vision.

What it solves

CV-CUDA is designed to eliminate bottlenecks in AI pipelines by providing high-throughput, low-latency image and video processing. It replaces slow CPU-based preprocessing up to the same GPU where the AI model resides, allowing for faster data flow and reduced costs.

How it works

It is a GPU-accelerated library of computer vision algorithms that integrates seamlessly with C/C++ and Python. It allows developers to perform operations like image resizing and decoding (via integration with nvImageCodec) directly on the GPU, avoiding the need to move data between the CPU and GPU memory.

Who it’s for

Developers building AI pipelines for computer vision on NVIDIA cloud, desktop, and edge platforms (including Jetson) who need to maximize processing speed and scalability.

Highlights

  • GPU-accelerated computer vision operators for high throughput.
  • Seamless integration with Python and C/C++ AI frameworks.
  • Support for a wide range of NVIDIA hardware architectures (Turing, Ampere, Ada Lovelace, Hopper, Blackwell).
  • Compatible with Linux, WSL2, and aarch64 platforms.

Related

  • Project
    NVIDIA/DALI

    A GPU-accelerated library for data loading and pre-processing that removes CPU bottlenecks in deep learning pipelines for image, video, and audio data.

  • Project
    inducer/pycuda

    PyCUDA is a Python wrapper for Nvidia's CUDA API that enables high-performance parallel computation on GPUs with automatic memory management and C++ speed.

  • Project
    NVIDIA-NeMo/Curator

    A GPU-accelerated data curation framework for building repeatable pipelines to load, filter, deduplicate, and transform massive text, image, video, and audio datasets for AI training.

  • Project
    NVIDIA/CUDALibrarySamples

    A collection of GPU-accelerated examples demonstrating the use of NVIDIA CUDA libraries for high-performance math, image processing, and linear algebra.

  • Project
    neka-nat/cupoch

    A GPU-accelerated library for rapid 3D data processing in robotics, providing high-performance point cloud algorithms, collision avoidance, and path planning.