It depends on application scenarios! Some (e.g. LLM) need bandwidth - other (e.g. Deep Learning AI) need, e.g., speed of graphics cores - yet other (e.g. large LLM models) also need sufficient volatile storage - other (e.g. office tasks) need none of that.
The interesting comparison is to the M5 Pro, since it has more memory bandwidth. Even the base M6 has twice the bandwidth of typical x86. Barely worth comparing until x86 has woken up after its decades-long sleep.
The latest Apple Mac mini retains its familiar chassis but delivers outstanding single-core and multi-core performance, thanks to the new M6 chip built using the 2NM process, that significantly outpaces the immediate competition from AMD, Intel, and Qualcomm.
English
Español
Français
Italiano
Nederlands
Polski
Português
Русский
Türkçe
Svenska
Chinese
Magyar