Vrm temps

[EcDarkSky]

Hello guys,

So the vrm is decreasing the voltage to the CPU voltage and turn the aditional voltage into heat so if you overclock your cpu that mean that CPU need more voltage so vrm don't have to decrease as much voltage so its coller. Conclusion overclocking CPU lower vrm voltage.

Am i right?

Just want to know.

Thanks for answer

[Mark77]

If you were talking about a linear VRM, then yes, your logic would make sense.  But the VRMs on computer motherboards for the past ~19 years or so have been "switch mode" VRMs (basically a technology similar to what's used in the big AC power supply in your computer!).  Which use pulse-width modulated switching of higher voltages to create lower voltages.  In a nutshell, they do not exhibit the same behaviour and the proportionality of power dissipation is to current (or more technically, the square of current, as in,  P = I^2R).  Since overclocking will induce higher current draw, obviously dissipation in such components will rise.

 

BTW, your message made me think of my old Intel P5 Pentium-60.  800nm technology (compared to 14nm today!).  Guess how much its motherboard's VRM dissipated?  *zero*, nothing, nada.  It ran the CPU directly on the main PSU's 5V rail!  But pretty much everything built after that, needed a VRM on the motherboard. 

 

 

[Deli]

You forget about current draw. Without overclocking, the CPU uses less power, and draws less current.

 

If your theory is correct, many Intel CPUs in idle state run at 800MHz, and 0.9V or less. The VRM will be hotter because it has to lower the voltage more. But it isn't the case.