[suggala]

Benchmark was executed only for CPU intensive node configurations. For highmem node configurations, AMD is not performing well. Probably need to title the report as Cloud Report for CPU intensive workloads. For memory intensive workloads, this report is not doing enough justice.

[Rafuino]

Can you elaborate on what you'd expect to see if they tested the highmem node configs?

[suggala]

Not just highmem configs. You have to do any decent memory intensive workloads to be able to see difference. Memory Access Latencies are higher for AMD machines because of various reasons.