PBO 5 3600
PBO 5 3600
As I've understood it... PBO Scalar relates to how long the VCore should be held elevated when the CPU is boosting; higher scalar values result in a longer hold-up.
If this is correct, I can see that it should be used judiciously. What I think happens is that while it can help the core hold a boost clock longer, it also generates more heat. The result being, as time goes on it should be less likely to boost to as high as the CPU warms more since the boosting algorithm sees less thermal headroom. Of course that's going to happen anyway but it could hasten the time...
As I've understood it... PBO Scalar relates to how long the VCore should be held elevated when the CPU is boosting; higher scalar values result in a longer hold-up.
If this is correct, I can see that it should be used judiciously. What I think happens is that while it can help the core hold a boost clock longer, it also generates more heat. The result being, as time goes on it should be less likely to boost to as high as the CPU warms more since the boosting algorithm sees less thermal headroom. Of course that's going to happen anyway but it could hasten the time until it starts reducing it's boosts.
But also, if your motherboard's CPU core voltage tends to have a lot of v-droop (poor LLC) then this could help because if voltage isn't high enough it also won't boost as high as needed to sustain the boost. In that case it would lower the boost frequency to maintain stability as the voltage sags some from the initial boost.
This is all conjecture of course. Nobody has solid concrete information on all this as AMD keeps this stuff very close... even theStilt hasn't talked about it that much. And also, so far anything PBO has very limited effect with Ryzen 3000 compared to the effect on Ryzen 2000. So it may actually have no effect.
Also, the CPU boost clock over-ride isn't reliable. On my board it actually makes my CPU boost less when I use it, if I set it to a high value (200Mhz) it will crash when I start any load. So be sure to test with and without to see what it does.
AMD isn't very open about details, and MB manufacturers are even less so. I'll see how things turn out when the new AGESA arrives next month. My BIOS is a bit confusing, with some duplicates and nothing else. It's not much better with other brands either. The promised boost clocks are now real, but will it perform better than that remains to be seen.