Hello,
I experiment a big performance difference for the same VM on 2 different ESXi servers (cloned from one to the other).
One is an old Dell Power Edge T610 (2 Xeon X5650), the other is a newer Dell Power Edge R720 (2 Xeon E5-2640).
The same program, which mainly does transformation (rotations and translations) of drawing primitives, uses 30% of the CPU on the VM of the old T610 and 60% of the CPU on the VM of the newer R720.
Both servers have nearly no activity. We have compared the ESXi configuration of the 2 servers they are almost identical.
I have also used the benchmarking tool CrystalMark 2004R3 on the 2 VMs, it shows better scores for the newer ESXi except for the GDI test which is better for the old one (about 10%).
How can we explain such a big difference? Can it be just because of the hardware or is there somewhere a hiden configuration parameter.