P40 vs K2

I run an Engineering team on VDI with 3 hosts that have K2s and just got a new server with a P40.
So far i am not experiencing any greatness with the P40. If anything i am seeing a performance decrease.

I am curious if others have seen the same thing?

K2 - 240Q profile
P40 - P40-8Q

servers are identical other than the cards.
running in HP GEN9 hosts

I also have considered running a few Engineers with a 1:1 GPU in our servers and disabling the FRL. i see massive gains when doing this on the K2s but nothing on the P40s does the P40 have FRL like the K2s did?

Any advice would be great!

New “performance lost” scheduler is active for Pascal card by default (see https://gridforums.nvidia.com/default/topic/743/#5282). Try to change scheduler (see https://docs.nvidia.com/grid/latest/grid-vgpu-user-guide/index.html#changing-vgpu-scheduling-policy).

There is no FRL active for Pascal.