Question High idle CPU core voltage ?

[FragTalentz]

Hi so recently i had to clear cmos and lost all my BIOS settings while redoing the settings on my 5600x i noticed that HW info was reporting a high idle CPU core voltage 1.325-1.425 while not doing anything at all really ( just HW info open ) , under load like cinebench R23 the core drops to 1.15-1.2 volts .
I have a negative offset per core -16 for my best -12 for my second best and -9 on the rest of the cores , thats what i tested as stable , i have DCOP on with 3666 on ram and FCLK on 1833 for a 1:1 and a +200 mhz boost to clock speed .
TDC- 90
EDC- 100
PPT- 85
These are settings i worked in for over a week , testing etc and has been stable for well over a year now . My problem is before i had to clear cmos my CPU core voltage would drop as low as 0.8-1.2 on idle now its pinned some 100-200mv higher and doesn't seem to fluctuate at all does anyone know why this is and if not is it safe to have that sort of CPU core voltage at idle ?
For all wondering all my testing was done with Cinebench r23, heaven for the GPU , Prime95 and Memtest oh also 3dmark for GPU , all stats/voltages etc where pulled from HW info as it's the one i trust the most .
The cpu voltage im looking at is also the CPU Core Voltage (sv12 TFN) .

 

[Lutfij]

Welcome to the forums, newcomer!

What is the make and model of your motherboard? BIOS version said motherboard at this moment of time?

 

[hotaru.hino]

Does it have high voltage when the clock speeds are at base clock speeds, or are one of the cores at turbo boost speeds? Because if any of the cores are at turbo boost speeds (which Windows defaults to an aggressive boost behavior, which causes the CPU to boost periodically to service housekeeping tasks), the voltage has to go up.

 

[FragTalentz]

Does it have high voltage when the clock speeds are at base clock speeds, or are one of the cores at turbo boost speeds? Because if any of the cores are at turbo boost speeds (which Windows defaults to an aggressive boost behavior, which causes the CPU to boost periodically to service housekeeping tasks), the voltage has to go up.

Highest i have seen the effective clocks are about 500 , the core clock (perf) has them jumping to 4.8 at times , im unsure which one to trust as martin says the effective clock is closer to what is actually happening .

 

[FragTalentz]

Welcome to the forums, newcomer!

What is the make and model of your motherboard? BIOS version said motherboard at this moment of time?

Mobo is the Asus strix B550-F gaming WI-FI and the bios version is 1401

 

[hotaru.hino]

Highest i have seen the effective clocks are about 500 , the core clock (perf) has them jumping to 4.8 at times , im unsure which one to trust as martin says the effective clock is closer to what is actually happening .

The perf core jumping to 4.8 GHz periodically is expected. As I mentioned, Windows still has house keeping tasks and it'll just shove it on one core, which causes it to boost.

Effective clock is just a measurement of clock speed including the periods where the core went to sleep. So if the core went to 4.0GHz for half a second and slept for the other, the effective clock speed is 2.0GHz (assuming a 1 second sampling period).

 

[FragTalentz]

The perf core jumping to 4.8 GHz periodically is expected. As I mentioned, Windows still has house keeping tasks and it'll just shove it on one core, which causes it to boost.

Effective clock is just a measurement of clock speed including the periods where the core went to sleep. So if the core went to 4.0GHz for half a second and slept for the other, the effective clock speed is 2.0GHz (assuming a 1 second sampling period).

Ok that makes sense but im seeing 500 AT MOST on effective clock and 4.8 on perf clock.

 

[hotaru.hino]

Ok that makes sense but im seeing 500 AT MOST on effective clock and 4.8 on perf clock.

Effective clock is expected to be really low in idle scenarios. If you run something heavy on the CPU, say Cinebench, then effective clock should be close to the reported clock speed.

The thing is when a core goes to sleep, its reported frequency stops updating. Hence why the need for the "effective clock" metric.

 

[FragTalentz]

Effective clock is expected to be really low in idle scenarios. If you run something heavy on the CPU, say Cinebench, then effective clock should be close to the reported clock speed.

The thing is when a core goes to sleep, its reported frequency stops updating. Hence why the need for the "effective clock" metric.

so should that result in basically a reported core voltage of 1.35.1.4 ? and is that safe for day to day use .

 

[hotaru.hino]

so should that result in basically a reported core voltage of 1.35.1.4 ? and is that safe for day to day use .

Yes. I should also add that basically all metrics stop updating when a core goes to sleep. And AMD processors are designed to go up to 1.5V.

 

[FragTalentz]

Yes. I should also add that basically all metrics stop updating when a core goes to sleep. And AMD processors are designed to go up to 1.5V.

well i never see above 1.415 v anyway , so thank you for the help and clarification . One last question if i may ? is it somewhat true that the volts dont really matter it's more temp ? for example under cinebench r23 30 min i dont go above 67c .

 

[hotaru.hino]

well i never see above 1.415 v anyway , so thank you for the help and clarification . One last question if i may ? is it somewhat true that the volts dont really matter it's more temp ? for example under cinebench r23 30 min i dont go above 67c .

As long as the part is running within spec, none of the factors really matter. AMD will guarantee at least 3 years out of the CPU. Beyond that, who knows how long it'll last?

 

[FragTalentz]

As long as the part is running within spec, none of the factors really matter. AMD will guarantee at least 3 years out of the CPU. Beyond that, who knows how long it'll last?

ty for the help.