Posts by Matthew Rocklin

Dask Survey 2021, early anecdotes

The annual Dask user survey is under way and currently accepting responses at dask.org/survey.

Read more ...


Stability of the Dask library

Dask is moving fast these days. Sometimes we break things as a result.

Read more ...


Dask User Summit 2021

Dask is organizing a user summit in mid-May. This will be a remote event focused on bringing together developers and users of Dask and the distributed PyData stack in different domains.

Read more ...


Estimating Users

People often ask me “How many people use Dask?”

Read more ...


Co-locating a Jupyter Server and Dask Scheduler

If you want, you can have Dask set up a Jupyter notebook server for you, co-located with the Dask scheduler. There are many ways to do this, but this blog post lists two.

Read more ...


Dask on HPC: a case study

Dask is deployed on traditional HPC machines with increasing frequency. In the past week I’ve personally helped four different groups get set up. This is a surprisingly individual process, because every HPC machine has its own idiosyncrasies. Each machine uses a job scheduler like SLURM/PBS/SGE/LSF/…, a network file system, and fast interconnect, but each of those sub-systems have slightly different policies on a machine-by-machine basis, which is where things get tricky.

Read more ...


Dask and ITK for large scale image analysis

Document headings start at H2, not H1 [myst.header]

Read more ...


Python and GPUs: A Status Update

This blogpost was delivered in talk form at the recent PASC 2019 conference. Slides for that talk are here.

Read more ...


Experiments in High Performance Networking with UCX and DGX

This post is about experimental and rapidly changing software. Code examples in this post should not be relied upon to work in the future.

Read more ...


Composing Dask Array with Numba Stencils

In this post we explore four array computing technologies, and how they work together to achieve powerful results.

Read more ...


Building GPU Groupby-Aggregations for Dask

Document headings start at H2, not H1 [myst.header]

Read more ...


Running Dask and MPI programs together

Document headings start at H2, not H1 [myst.header]

Read more ...


Single-Node Multi-GPU Dataframe Joins

Document headings start at H2, not H1 [myst.header]

Read more ...


Dask, Pandas, and GPUs: first steps

Document headings start at H2, not H1 [myst.header]

Read more ...


GPU Dask Arrays, first steps

The following code creates and manipulates 2 TB of randomly generated data.

Read more ...