GPU Data Science with RAPIDS

Accelerate common data science workloads on GPUs

Here are some example notebooks that walk through using a GPU with RAPIDS, scaling to a cluster with Dask, then some runtime comparisons. To quickly run these notebooks in Saturn Cloud, use the RAPIDS resource template to create a premade resource. This will create a resource with notebooks automatically loaded into the Jupyter server:

Rapids template resource

Use RAPIDS on a single GPU

Minimal changes required from familiar pandas and scikit-learn code

Use RAPIDS on a GPU cluster

Scale to larger data sizes with multiple GPUs

Need help, or have more questions? Contact us at:

We'll be happy to help you and answer your questions!