CPU overclocking:
Our CPU overclocking part will focus on the performance of the i9-12900KS. It is important to underline that this Biostar Z690A Valkyrie is not intended for advanced overclocking. But we’ll see that it does quite well despite the fact that it is not in the best conditions for this exercise.
Protocol:
Before starting, it is necessary to have one or more reference scores with our i9-12900KS at stock frequencies , that is to say, without having made any modification except having activated the XMP profile in the case of an Intel processor. Concerning the test voltages, we decided to put the voltage in fixed mode.
It is also very important to have a sheet of paper or a spreadsheet at hand, which will allow you to note all the tests you are going to perform as well as the results or failures.
Load-Line Calibration:
Before starting, we need to set the LLC. In the case of this motherboard, we set it to level 1. Although this is an important point concerning stability, here only mode 1 was viable with an identical voltage in load and idle. The other modes showed excessively high Vdroop.
Benchmark Scores:
So let’s start by noting the results under Cinebench R15 and R20, taking care to note the voltage(vcore), the max temperature during the benchmark, the power consumption and of course, the score obtained. During the benchmark, the HWinfo window is open which has a slight negative impact on the score.
Overclocking under watercooling in stepsapes :
To start with, I leave the BIOS in auto by just setting the RAM to 3200C14. First, we read the vcore voltage and load frequency during the benchmarks on Cinebench R15 and R20 (5200 at 1.32v). I have chosen a starting frequency of 5200 MHz with a multiplier of 52, a BCLK of 100 and a voltage of 1.2 volt. The Ecore and the cache are placed at 4000 MHz. The idea is then, in the OS, to test the stability on several runs of Cinebench R15 and R20 multithread.
If it’s stable, I increase the frequency by 100 MHz while checking the stability again. If the benchmark crashes, I increase the voltage by 0.05 volt steps in order to find a stability allowing to run the benchmark. The objective is to get a “quick” idea of the maximum benchmarkable frequency.
Note : point to underline, the cache remained at 3600 MHz whatever the requested value when the Ecore were active.
Thus, my 12900KS managed to reach 5.4 GHz with a voltage of 1.3 volt, but with a very high temperature. As for the Vrm’s, they reached 45°C at their hottest. After that we mounted the cache and the Ecore which required more voltage. And finally we tested this overclocking without the Ecore.
And here are the results on R20. As you can see, 5400 MHz is close to the temperature limit. For a normal use, I would thus opt for a frequency of 5300 MHz.
So here are the two screenshots of our tests with Ecore with the i9-12900KS.
And, those without Ecore.
