Check if your 9900K is experiencing overvoltage conditions.
Check if your 9900K is experiencing overvoltage conditions.
Hey everyone. I've successfully achieved stable performance at 5200MHz on my 9900K stable. I've tested everything—Cinebench R15, R20, XTU, AIDA64, CPU-Z Bench, 3DMark, and even played several games for hours—without any blue screens. My voltages and LLC settings are set as follows: Adaptive Mode with a turbo vcore of 1.150v, offset 0.210, resulting in a vcore of 1.360. I'm using an ASUS Maximus XI Hero Z390 board. When I run Cinebench while monitoring core temperature, my vcore spikes to 1.50v. Is this normal? Should I worry about these readings? Or are they accurate? Thanks!
Offset is off by a bit. 1.50v isn't correct. The numbers should be checked against the VID. Make sure it's not set lower than the VID, or it won't function. It seems like it's applying an extra 200mV plus some value. You might want to lower that offset before risking damage. How did you manage to cool it at such a high temperature?
I've been looking into the 9900K's voltage and VID specs. I adjusted my settings to: Adaptive Mode, Extra turbo vcore @ 1.250v, Offset @ 0.110 = 1.360 vcore for the LLC6. Those seem fine with what I see. Any tips for improvement? I'm using a Corsair Hydro Series H115i Platinum RGB with Thermal Grizzly Conductonaut Liquid Metal Paste. Room temps are low—around 10°C, and it'll drop even more in winter. I’ll need a coat soon. Idle temps range 23°C–27°C across 8 cores; gaming temps 55°C–68°C; stress temps 80°C–95°C (sometimes up to 100°C). The stress tests and the heat they generate don’t worry me, since I’m mainly a gamer. What about overvoltage of 1.5V occasionally? Is that safe for my CPU?
You installed HWiNFO64 and ran a quick CPU-Z benchmark, which shows the maximum vcore as 1.40–1.41v. That’s significantly higher than what you saw in CoreTemp (1.5v). It makes sense—CPU-Z often gives a more accurate measurement of real-world performance. The difference could be due to how each tool samples or processes data. No need to worry, it’s normal for the two to vary.
Coretemp is an older tool with some updates. It has its advantages but lacks CPU Vcore display. Instead of VID it shows VID, which isn't directly usable as VCORE unless specific conditions are met. Matching VID and VRM loadline calibration can be confusing for average users, as Loadline Calibration is a resistance measurement (see the referenced link). Even if they match "DC Loadline" to your VRM, it doesn’t guarantee VID syncs properly with Vcore. Plus, Coretemp may attempt to install unwanted software. It’s mainly helpful for temperature readings. With HWinfo64 you can connect to Rivatuner Statistics Server to monitor temps during games—so why rely on Coretemp? Your temperature offset remains high. Consider adjusting it: try Additional Turbo Voltage at 1.35v, remove the offset gradually (starting from 0.01v), and check stability.
You're seeing blue screens during startup because the system is struggling with the settings. Applying an offset didn't improve the issue. Your current settings seem to be working well now. Higher offsets might help in some cases, but it depends on your specific hardware and configuration. Test different values cautiously.
I don’t know your default VID. Offsets depend on your base VID. The "AC Loadline" affects your base VID too, but the AC LL adjustment (possibly linked to SVID behavior) changes this only for advanced users. I assumed a 1.35V boost would provide at least 1.35V for the CPU. It seems the system crashed unexpectedly. Without an Asus board, I can’t offer reliable advice. You might try 1.35V boost plus +50mV?