Standard deviation of CPI: A new metric to evaluate architectural time predictability

This paper first proposes the concept of architectural time predictability, which separates the time variation due to hardware architectural/microarchitectural design from that caused by software. We then propose a new metric - standard deviation of clock cycles per instruction (CPI), to measure architectural time predictability. Our experiments confirm that standard deviation of CPI is an effective metric to evaluate and compare architectural time predictability for different processors.