dask-cuda
Utilities for Dask and CUDA interactions
Decision gist · record as of 2026-08-14
Yes, if you need to deploy Dask across multiple GPUs. The library is actively maintained, has no known vulnerabilities, and uses a permissive license. Install friction is moderate due to CUDA/driver dependencies, but those are prerequisites for any GPU work anyway. Not suitable if you lack GPU hardware or are looking for automatic GPU code conversion.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires CUDA-enabled GPU hardware, NVIDIA drivers, and CUDA toolkit installed on the system; Python >= 3.11.
- Medium install friction due to dependencies on cuda-core and nvidia-ml-py, which require CUDA toolkit and NVIDIA drivers to be present on the system.
- Actively maintained with recent releases; last commit 2026-08-13.
License · maintenance · safety
Apache-2.0 (permissive) — Apache-2.0 permissive license allows commercial and private use with minimal restrictions, making it suitable for most deployment scenarios.
last release 2026-08-06 (8 days) · last repo commit 2026-08-13 · 324 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 302,340 downloads/mo, #7,824 on PyPI
Alternatives
Verify before relying
pip install dask-cuda
from dask_cuda import LocalCUDACluster
from dask.distributed import Client
cluster = LocalCUDACluster()
client = Client(cluster)- Whether the package automatically handles GPU memory management and spilling between GPUs and host memory.
- Performance characteristics and scalability limits for typical multi-GPU cluster sizes.
- Compatibility with specific RAPIDS libraries (e.g., cuDF, cuML) beyond the documented integration points.
What it is and what it does
Dask CUDA is a utility library that handles the deployment and lifecycle management of Dask workers on systems with multiple CUDA-enabled GPUs. It provides a LocalCUDACluster abstraction that simplifies spinning up a distributed Dask cluster where each worker is bound to a specific GPU, allowing you to parallelize work across available hardware without manually managing worker-to-GPU assignment.
The library does not automatically convert existing Dask code to run on GPUs—that remains the responsibility of your application code and the libraries you use (such as RAPIDS or CuPy). Instead, it focuses on the infrastructure layer: making it straightforward to create and manage a multi-GPU Dask cluster, handle worker placement, and coordinate resource allocation. It depends on click, numpy, pandas, zict, nvidia-ml-py for GPU monitoring, and rapids-dask-dependency for integration with the RAPIDS ecosystem.
Use it for
- Set up a multi-GPU Dask cluster on a single machine or cluster for distributed data processing with GPU acceleration.
- Manage worker-to-GPU affinity automatically when scaling Dask workloads across available CUDA devices.
- Integrate GPU-accelerated RAPIDS libraries (cuDF, cuML) with Dask's distributed scheduler for large-scale data science.
- Monitor and coordinate GPU resource allocation across multiple Dask workers in a heterogeneous cluster.
- Prototype and deploy distributed GPU workloads without manually writing worker spawning and GPU binding logic.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you need to deploy Dask across multiple GPUs.
The library is actively maintained, has no known vulnerabilities, and uses a permissive license. Install friction is moderate due to CUDA/driver dependencies, but those are prerequisites for any GPU work anyway. Not suitable if you lack GPU hardware or are looking for automatic GPU code conversion.
Install
dask-cuda on PyPI
Before you install
Medium install friction due to dependencies on cuda-core and nvidia-ml-py, which require CUDA toolkit and NVIDIA drivers to be present on the system. Actively maintained with recent releases; last commit 2026-08-13.
Requires CUDA-enabled GPU hardware, NVIDIA drivers, and CUDA toolkit installed on the system; Python >= 3.11.
License in practice
Apache-2.0 permissive license allows commercial and private use with minimal restrictions, making it suitable for most deployment scenarios.
Quickstart
pip install dask-cuda
from dask_cuda import LocalCUDACluster
from dask.distributed import Client
cluster = LocalCUDACluster()
client = Client(cluster)Verify before relying
- Whether the package automatically handles GPU memory management and spilling between GPUs and host memory.
- Performance characteristics and scalability limits for typical multi-GPU cluster sizes.
- Compatibility with specific RAPIDS libraries (e.g., cuDF, cuML) beyond the documented integration points.
Package facts
| License | Apache-2.0 permissive |
| Python support | Supports the current Python release >=3.11 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 7 packagesclickcuda-corenumpynvidia-ml-pypandasrapids-dask-dependencyzict |
| Maintenance | Actively maintained 8 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 302,340 / month, #7,824 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Intended Audience :: DevelopersProgramming Language :: Python :: 3Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Topic :: DatabaseTopic :: Scientific/Engineering |
Evidence: dask_cuda-26.8.0-py3-none-manylinux_2_28_aarch64.manylinux_2_28_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “dask gpu cluster management”
- dask-cudaDask CUDA provides utilities for deploying and managing Dask workers…
- distributed-ucxx-cu12Provides a high-performance UCX communication backend for Dask…
- dask-cudf-cu12Dask cuDF extends Dask DataFrame with a GPU-accelerated backend,…
Give your agent the search over MCP, or paste the wish link into any chat.
More Scientific/Engineering packages
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
pandas provides fast, flexible data structures (Series and DataFrame) for loading, cleaning, transforming, and analyzing labeled or relational data in Python.
scipy provides numerical algorithms for mathematics, science, and engineering—including optimization, integration, linear algebra, Fourier transforms, signal and image processing, and ODE solvers—built on numpy arrays.
scikit-learn provides a comprehensive Python library for supervised and unsupervised machine learning, including classification, regression, clustering, dimensionality reduction, and model evaluation tools built on NumPy and SciPy.
Install it if you need to train, evaluate, or deploy supervised or unsupervised learning models.
dill extends Python's pickle module to serialize and deserialize a much wider range of Python objects, including functions, lambdas, classes, and interpreter sessions, to byte streams for storage or network transmission.
Multiprocess is an enhanced fork of Python's standard multiprocessing library that uses dill for better serialization, allowing you to spawn processes with a threading-like API and share complex objects between them.
Install it if you use multiprocessing and encounter pickle serialization limits with lambdas or complex objects.
See also dask-cudf-cu12 · rapids-dask-dependency · raft-dask-cu12 · dask-jobqueue · coiled · nvidia-cublas · nvidia-cuda-cccl · nvidia-cuda-runtime · nvidia-cuda-runtime-cu11 · libraft-cu12