Benchmark scores are comparable between different hypervisors (vmware, citrix) and OS

[owiegmann]

We make some tests in virtual machines that are running on different hypervisors. We get on the same physical hardware very different results for memory scores.

VMware: 700
Citrix: 1200
Physical: 5000

How valid is it to compare the results from the different platforms?

[David (PassMark)]

Yes, PerformanceTest can run on VMs.
I am not sure if the numbers you posted are the Passmark rating or some other number. Physical machines will always be faster however.

If the VM doesn't fully emulate a physical machine then you can get some low scores. For example the overall PassMark rating takes into account 2D and 3D video performance. If the VM doesn't support 3D at all, then the overall score will be low because you won't be able to run those tests.

Depending on what you are doing with the VM the overall PassMark rating might not be the best performance measure. For example if the machine is going to be a web server, then you should focus on the disk and CPU performance, but ignore the 2D and 3D performance.