I'd say this usage is very specific - that the best way to move forward would be to run the workload on real hardware. You might get a huge benefit from a high-core count in certain projects while other codebases will build faster on fewer cores with higher single-thread performance.
> the best way to move forward would be to run the workload on real hardware.
OP is asking whether the hardware will perform so he know whether to buy it. Is the "best way to move forward" really to buy the hardware and find out?
