Yes, that's pretty exciting! There is even support in Polars to execute the dataframe compute in DataFusion (as it can handle larger than memory datasets).