3 ms·

Then once the subset of data is in distributed memory... you hittem w pyspark and all compatible libraries.