11 Jan
2018
11 Jan
'18
6:05 p.m.
Hi there, Are you sure there is no bug in how "new" binary bulk load is loading data in parallel?
From my tests it was at least 5x speedup [when loading data from TXT to MonetDB] with respect to binary bulk load.
And I observed that only 1 CPU was used during BINARY bulk load. And ~16 CPUs (all available) were used during CSV loading. Thanks, Anton