Comment by gatesn

How push-down compute helps Vortex run faster and decompress less