Why is there almost none published overclocking tests for AM5 motherboards?
Ryzen 7 9800X3D which runs on AM5 socket is best for gaming, and yet I only have 1 article which shows results of testing some AM5 motherboards.
See here for "DDR5 Maximum frequency" and "Operating Frequency".
The motherboard that can extract more juice frequency out of RAM or CPU, is the one what was designed better considering the following:
More extra layers on PCB, you can split and put the traces for one channel on a few layers, traces from the other channel on other layers, have a ground plane or something between traces for extra shielding, etc
Ram topology complex routing method.
more phases in the VRM
add a multiphase power with 2-3x the typical phases
glass mat/weave orientation
As for CPU cure clock, it is mostly due to power integrity differences; modern CPUs generate very fast current draw transients (~1010 A/s) which causes voltage to momentarily spike down; these voltage minima are what limit stability. However, thermal/degredation issues come from average voltage. A better decoupling setup and better buck controller/firmware and phase design can mean 100-200mV improvements in min voltage (at least at board level) compared to low end boards.
I'm curious to see how would Asus ROG X870e Apex fare against GIGABYTE X870E AORUS MASTER or MSI MAG X870 Tomahawk WiFi.