We recently ran a benchmark comparing Bodo and Dask for distributed data processing in Python. In this particular test, Bodo showed significant speedup (~50x) over Dask, and we believe this is due to our auto-parallelizing compiler and MPI-based backend. We encourage others to reproduce our results and let us know how we can improve our benchmarks to make them more informative, useful, etc. Check out the full write-up here, and our GitHub repo here.
There is also a discussion about this benchmark in Github Issues here which provides additional context that is useful for understanding the results.
Hi @Scott_Routledge, welcome to Dask Discourse forum,
I see you already found the discussion between some of the Dask maintainers, but thanks for sharing this benchmark here.