rapidsai/dask-cuda

Utilities for Dask and CUDA interactions

What it solves

It simplifies the deployment and management of Dask workers on systems equipped with CUDA-enabled GPUs, making it easier to orchestrate GPU-accelerated computing tasks.

How it works

It provides utilities, such as the LocalCUDACluster class, to help users set up and manage clusters of Dask workers specifically optimized for CUDA environments.

Who it’s for

Developers and data scientists who are using Dask for parallel computing on multi-GPU systems and utilizing GPU-accelerated libraries like RAPIDS or CuPy.

Highlights

  • Provides specialized utilities for CUDA-enabled systems.
  • Simplifies the creation of CUDA clusters via LocalCUDACluster.
  • Designed to work alongside GPU libraries like RAPIDS and CuPy.

Related

  • Project
    NVIDIA/cccl

    A unified collection of core CUDA C++ libraries (Thrust, CUB, and libcudacxx) providing high-performance parallel algorithms and standard library abstractions for GPU development.

  • Project
    NVIDIA/cudf

    A GPU-accelerated DataFrame library for tabular data processing that provides a pandas-compatible API and zero-code change acceleration for pandas, Polars, and Dask.

  • Project
    kunzmi/managedCuda

    A .NET wrapper for NVIDIA's CUDA Driver API and libraries, enabling C# and Visual Basic developers to integrate GPU acceleration into their applications.

  • Project
    lattice/quda

    A GPU-accelerated library for lattice QCD calculations, providing optimized Dirac operators and solvers for various fermion actions on NVIDIA hardware.

  • 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.