I would like to use spark with Rapids. however, I didn’t found guide for how to measure on a single node. when I try to measure it on EMR using aws guide here:
I’m getting low performance (0.1x compare to similar instance on x86). how can I measure on Rapids single node and get significant speedup?