HP v.s. Dell which spec is better

really, this is the RAM CAS? I did not check that, my bad.
If so I would agree with you andyalder, specially for the RAM side, which is really a major hardware need for a vmware server, these 2 specs cannot be compared.
With 16GB RAM ESXi server nowadays you will do... well, about nothing!
With a 256GB RAM server, depending on other specs (storage system performance  and size for example), you could have dozens of VM.

HP will charge for firmware downloads when warranty ends.
DRAC Enterprise is fully functional, to get ilo advanced "add 300$"

Ok if I up the HP to 14x16GB which model would you pick if you had to have a VMware test box?
What is the difference now just processor?

Depends on workload.
2x12-core CPUs will fit 3x more virtual machines 1x8core can fit.
Each of them is overkill for a test machine (test VMs would be 2-4GB each, so you can stuff like 100)
First you will run short is RAM no matter how much you put.
If you pay electricity consider low power CPU and low power RAM (only I see low power RAM in your HP)
Calculation performance calculator here: http://www.novatte.com/our-blog/197-how-to-calculate-peak-theoretical-performance-of-a-cpu-based-hpc-system
Another plus for DELL - you can fit any compatible RAM there and keep warranty on DELL components.

We want to be able to kick up VMs to test products with various configurations to match client environments
We will likely be running 10-20 VMs

So you would lean HP from comment?

Nope, I lean towards AMD.
10-20 VMs equals 40 vCPUs, which will have tough time on 8 cores, so better off with 24 cores. Basically more than 2 vCPUs per core is a crowd once they get used (hyperthread is not separate core)

Hi,
I think that in your initial question the cpu part is as the RAM part, you cannot compare it.
I did not see that too, that the HP config was with only one physical processor.
If you compare 2 cpu E5-2670 with the AMD one then yes the AMD will have more cores but less frequency.
This is a choice and it depends, but considering you don't want to create a lot of VM personnaly I would rather choose frequency over cores number.
Also in general for a test environnement it is far enough.. for both configuration Dell and HP.
And I really agree that HP servers are more expensive when we compare two configurations with same specs.

Where will you put that frequency? I'd prefer low power option with most cores...

it would be more cpu powerfull for the VM used in the test environment.
It is okay not to agree.. : )  maybe you're right.
It depends what is really the purpose of the test environment.

You cannot compare Xeon cores with Opteron cores on a one to one basis. Instead you should compare a Xeon based machine with an Opteron based one under VMmark tests. Try to pick a single manufacturer and find one of each server such as comparing the DL380p with the DL385p - http://www.vmware.com/a/vmmark/1/submitter/Hewlett-Packard/ (Note though that they put more RAM in the DL380). For a Dell version of the comparison select submitter=AMD to see the Opteron based Dell R715. You may find the AMD CPUs are a bit cheaper though and as gheist says you're unlikely to be bottlenecked on CPU power anyway.

BTW, there's a DL380 Gen9 in the VMmark test results so if you hold off a bit the Gen8s will no doubt drop in price, Dell probably have new models in the pipeline too.

Great points!  Thanks.