For the past three years, I had my CPU overclocked to 5GHz, and it ran flawlessly. When idle, temperatures hovered around 45°C, and under maximum load, they typically peaked at 80°C, occasionally spiking to 90°C but only briefly. Under normal load, the temps were around 60°C. However, this summer, I began experiencing rare BSODs while using Adobe Lightroom Classic under heavy load (which is the most intensive task my system handles, even more than gaming). These crashes seemed to occur when the CPU temperatures hit 90°C, and I initially suspected the unusually hot summer temperatures were to blame.
Concerned, I decided to reset the BIOS to its optimized defaults. Now, the idle temps have dropped slightly to around 42°C, but under the same heavy Lightroom Classic workload, temperatures on some cores are hitting 95°C, with one or two cores even briefly touching 100°C. For example, if the operation lasts for 1 minute, the average temp stays around 85°C, occasionally dropping to 65-70°C, but still spiking to dangerous levels (95°C to 100°C). This is particularly alarming because it's happening on the so-called "optimized defaults."
Shouldn’t there be safeguards in place to throttle the CPU when it hits these temperatures? Oddly, while my old overclock was set to 5GHz across all cores, the optimized defaults are now pushing most cores to 5.3GHz! I’m more concerned about the high temperatures than the clock speeds at this point.
Could this be an issue with cooling or thermal paste? I’m not an expert in BIOS settings or overclocking, but I’d like to know if there’s something I can adjust in the BIOS to address the temperature issue?
System Specifications:
- Motherboard: ASUS ROG STRIX Z490-G GAMING
- CPU: Intel Core i9-10900K
- RAM: ADATA XPG SPECTRIX D50 RGB 32GB DDR4-3600 CL18
- Cooling: SILVERSTONE Permafrost 240mm ARGB AIO
- GPU: Nvidia RTX 2080Ti
- Case: Fractal Design Define 7 – Black Solid
- PSU: SilverStone 850W 80+ Gold Full Modular
What could be causing these issues, and is there anything I can do to keep my CPU temps in check?
2 Comments
repaste and reseat your cooler.
there is a slight chance that the pump in it is not working, or you have the water pump plugged into something weird.
plug water pump into either aio pin header or cpu, then set to 100% all the time in bios. Its ok if your pump goes full throttle all the time.
you could try running a continious benchmark and then see if the rad is heating up. if cpu temps go high but rad is not hot at all then the pump shat itself. rma / warranty / buy a new one.
tjmax is 100C on a 10900K, that is when it starts to throttle. Spiking to that level won’t hurt the chip, it will just slow down to maintain that number.
“Optimized Defaults” probably means Asus is fucking with the power limits and enabling multi-core enhancement. A stock 10900K should hit 5.3GHz but not on all cores. Welcome to the game of “defaults” where Intel plays chicken with the motherboard vendors.
https://ark.intel.com/content/www/us/en/ark/products/199332/intel-core-i9-10900k-processor-20m-cache-up-to-5-30-ghz.html
You probably need to replace the thermal paste. Pump-out affect is when heat cycles force thermal paste away from the die of the CPU. Low viscosity thermal pastes usually perform better initially but become worse over time.