> In fact for some specific power targets (mobile for example) they are more efficient than the corresponding Intel CPUs.
I agree, but only if the measurement of power efficiency includes idle periods -- eg. "browse for 5 minutes, lock phone and wait 30 minutes, repeat".
I don't believe you can provide a source showing an ARM CPU executing a particular workload (i.e. not including idle periods) at greater power efficiency than a same-generation Intel x86 CPU.
I agree, but only if the measurement of power efficiency includes idle periods -- eg. "browse for 5 minutes, lock phone and wait 30 minutes, repeat".
I don't believe you can provide a source showing an ARM CPU executing a particular workload (i.e. not including idle periods) at greater power efficiency than a same-generation Intel x86 CPU.