Reporting Deviation in Power Measurement (Precision)
Reporting Deviation in Power Measurement (Precision)
Check if the numbers fit expected ranges. If they don't, adjust or investigate further.
The HWInfo tool has launched a beta update featuring Power Reporting, sensor CPU lifespan improvements, and MOBI enhancements.
It seems your MSI board is in good shape. The main upgrade would be replacing it with a 300$ motherboard to save the cost of your 200$ CPU. I’m leaning toward Intel next time, especially since AMD’s deals don’t seem to work for me. (They do allow it, apparently.)
Discussions in OC forums have centered around this topic, but it's essentially a low-level BIOS adjustment unrelated to AMD. It seems the blame often falls on those who create the BIOS files for these boards. To test different revisions, you can compare performance under full load, noting that accuracy is limited to around 100% and typically varies by about 5%. A range of 95 to 105% is generally acceptable.
Just mentioning this could mean AMD should reconsider their licensing approach. From their side, it would feel like a direct hit, but that doesn’t excuse the issue. PS: I just verified it’s 100% at full load—his explanation was clear and logical. Previously, the numbers varied widely (like 500%) and didn’t make sense to me.
This situation is largely beyond my control. AMD might not be aware or concerned about it. If it leads to real CPU failures, that would change things. These reports come from software dealing with algorithms instead of direct measurements. The temperature and voltage changes we observe occurred roughly a thousand milliseconds ago, meaning they’re not happening in real time. For instance, when a CPU hits its thermal cutoff and the board stops, there’s about a 10% chance (at most) of actually detecting the shutdown temperature.