8600k with 5GHz on Asrock Z370 Extreme4 – any power adjustments?
8600k with 5GHz on Asrock Z370 Extreme4 – any power adjustments?
Hello, after experimenting with my 8600k I’ve successfully hit around 5GHz while maintaining stable temperatures and voltages (about 80°C and below 1.350V). However, I’m experiencing significant thermal throttling, which causes the CPU to slow down to near 3.6GHz. The HWinfo64 data from the 15-minute Prime95 FFT run shows this issue clearly. My current setup includes a core ratio of 50%, cache ratio 45%, PL1 at 120, PL2 at 150, offset 140mV, and LLC2 enabled. I suspect the power limits I’ve configured are restricting the clock when power draw is too high. One possible fix would be to raise the CPU thermal throttle temperature cap to 90°C and then adjust PL1 based on actual power consumption. But I can’t find that adjustment in my BIOS. I’m open to suggestions—what you think about this approach? Is it feasible with what I have? Thanks a lot!
I attempted to increase the settings to PL1 190 and PL2 210, but temperatures hit 94ºC within five seconds after launching Prime95. Using AUTO didn’t help either, as the entire HWinfo turned red immediately upon starting the test.
It seems like things are behaving oddly now. Back then with an 8600K, I ran at 5GHz around 1.38v using a similar or better cooler—probably a Cryorig H7. The board used a Prime Z370-A, and you're right about the mounting pressure; it was actually maintaining the set voltage under load. Some boards have very aggressive LLC settings, and ASUS tends to enable a lot of that by default, which can cause the voltage to spike beyond what you set in BIOS. My 8600K wasn't overclocked and performed well—it had just six cores with no hyperthreading.
Intel's TIM is favored for long-term stability. It’s expected to remain functional... likely for many years. However, its performance isn’t ideal, so high-temperature processors often replace it. An 8600K shouldn’t become noticeably hotter over a short period like five years, though no one has confirmed this yet.
It's been over four years since I last changed the thermal paste. I'm planning to try it again before removing the heatsink, though I'm not sure it will make much of a difference. I'm pretty sure the Noctua cooler is set up correctly. Would using a fixed Vcore instead of offset be useful? Also, I've seen advice suggesting avoiding LLC1 and keeping the Vdroop low—really important for stability.
I suspect I may have damaged my PC! Last night I noticed my voltage rising to 1.37/1.38v while playing with the previous stable configuration I used (the exact same settings I shared in my original post, with Prime95 maxing at 1.344v and temperature reaching 79/81°C under load). Those settings were saved in one of my BIOS profiles, and I usually revert to that profile whenever I felt unsatisfied with the changes. After researching situations where new BIOS updates didn’t update correctly (thinking it was due to rising voltages during testing), I chose to reset to the default UEFI settings and restart the system to start fresh. The result is a complete lack of the boot screen, no windows appearing,... nothing. I’ve tried various fixes: connecting multiple monitors via HDMI, DP, and VGA, swapping the GPU, replacing the PSU, disconnecting everything from the motherboard except CPU and RAM, removing RAM one stick at a time, and even clearing CMOS by removing the battery and adjusting jumpers per the manual. I’m unsure if I’ve fried my CPU, destroyed the motherboard, or if there’s another issue I can’t identify. Honestly, yesterday was my birthday and it didn’t go as planned when I opened my eyes. Any suggestions?