Hacker News Reader

If you want built-in GPU support (and distributed), you should check out cuNumeric (released by NVIDIA in the last week or so). Also avoids needing to manually specify chunk sizes, like it says in a sibling comment.

https://github.com/nv-legate/cunumeric

View on HN

I see they also have have pandas replacement: https://github.com/nv-legate/legate.pandas. How is it different from cuDF?

I believe cuDF is not distributed? If you use Legate Pandas (or any Legate libraries) you'll get distributed as well as GPU execution out of the box.

You can probably use https://github.com/rapidsai/cudf/tree/main/python/dask_cudf a dask wrapper around cuDF.

Reply on news.ycombinator.com