Does this have relation to Windows DPC Latency such as the Invariant TSC's & Dynamic Tick feature with UsePlateFormClock?
MissFortuneGO argument is right! Evidence would be different if they had used a precision laboratory clock! UPDATE: However, further reading indicate that the occurring problem is real!
See no big deal here, as i seldomly use benchmarks which even in case of synthetic ones don´t have much significance. After ugdate of graphic drivers i test with a hardware hogging game like 'Metro' or 'Battlefield 3'.
Benchmarks aren't the only things affected by an off RTC though. If the timing of the entire OS is off, it could lead to some weird (and probably unnoticeable) stuff. I recall quite a few people having issues in the past with World of Warcraft where cooldowns on spells would finish/be delayed quicker/later than they were on the server due to core syncing with multi-core CPU's (not 100% sure if WoW relied on RTC or not though) Can only imagine framerate, input latency, frame time, audio, and other sensitive stuff would be affected by a non-accurate clock. Over-zealous anti-cheat engines may probably flag you too. Then there's tiny stuff, like your system clock itself just being off
I get the feeling this is more of a non hardware pure hpet + dynamic clocks. since this issue is more aparent when OC/UC which is an hpet issue... i NEVER had timer issues in windows8 since RTM leaked. did anyone bother to test if this 2 things were being used for the RTC in windows8? I been running with this and my benchmarks has always been on par with my GFs win7 pc, never had anticheat issues and never noticed my clock being off : bcdedit /set useplatformclock true bcdedit /set disabledynamictick yes