I know everyone recommends stress testing one's OC configuration with some synthetic application like prime95, AIDA64 or whatever, including the savvy guys. But I'd like to actually learn on the actual point of it.
Fact:
The stress/temperature imposed during synthetic testing is unachievable in many real scenarios (especially gaming!*).
*ArmA 3 example: Cores 1-3 50-80% utilization, Cores 4-5 ~30% utilization, Core 6 almost not utilized, plus most probably only partial utilization of CPU feature sets.
Problem:
So why would I bother hammering my CPU under conditions it will never experience?
The only thing that I could think of is:
'If the particular OC config gives errors in Prime95 it means that something bad is going on in the CPU and this bad thing will continue to be going on at that OC config even if CPU load is at 60-80%.'
Disclaimer:
I OC the rig exclusively to play ArmA. Once I finish my session I load up a power-saving oriented BIOS profile.
Fact:
The stress/temperature imposed during synthetic testing is unachievable in many real scenarios (especially gaming!*).
*ArmA 3 example: Cores 1-3 50-80% utilization, Cores 4-5 ~30% utilization, Core 6 almost not utilized, plus most probably only partial utilization of CPU feature sets.
Problem:
So why would I bother hammering my CPU under conditions it will never experience?
The only thing that I could think of is:
'If the particular OC config gives errors in Prime95 it means that something bad is going on in the CPU and this bad thing will continue to be going on at that OC config even if CPU load is at 60-80%.'
Disclaimer:
I OC the rig exclusively to play ArmA. Once I finish my session I load up a power-saving oriented BIOS profile.
Last edited: