High idle temperature issue following MOBO update
High idle temperature issue following MOBO update
Hello, I’m facing an unusual problem with my hands. After trying everything I could on my own, I’m hoping for some guidance to identify the cause. Here’s what happened – I switched from a ROG X670E Crosshair Hero to an MSi X870E Edge Ti. The main issue started after Asus began having recurring stability problems: BSOs, crashes, black screens, audio issues, and more. I was managing these troubleshooting sessions at least twice weekly.
Once the switch happened, everything felt much smoother – solid performance with no glitches. The only exception was temperature. On my old board, idle temps stayed around 45°C, while on this new setup they hovered near 60°C and reached up to 80°C during boot. Ambient conditions remained unchanged. The AIO is connected to the CPU fan header.
I’ve tried several fixes:
- Reapplied thermal paste multiple times in different patterns (MX-6 and Kryonaut) – no improvement.
- Changed mounting positions on the AIO – both top and front, but temperatures stayed the same.
- Switched to a Thermal Grizzly contact frame for the MOBO and swapped back – still no change.
- Inspected the AIO pump mount to ensure it wasn’t too tight or loose.
- Experimented with Ryzen Master, set control mode to Eco – temperatures didn’t shift.
- Tried BIOS tweaks (X3D Gaming, PBO), Windows settings, power plans – nothing altered temps.
- Updated BIOS and drivers, removed Armoury Crate from Asus.
- Double-checked PSU connections (two EPS 8 pin cables, same as before).
What I didn’t do:
1. Perform a full Windows clean install (last done ~1 week ago).
2. Replace the cooling solution (no other coolers available).
3. Clean the system thoroughly.
Potential causes for the sudden temperature rise could include:
- Improper thermal paste application or contamination.
- Incorrect AIO mounting or fan alignment.
- Dust buildup affecting airflow.
- A failing or loose cooling fan.
- Unexpected power delivery issues.
To bring things back under control, I recommend:
- Verify all fans are properly seated and functioning.
- Ensure the AIO is correctly mounted and not obstructed.
- Check for dust in vents and clean if necessary.
- Monitor temperatures closely with a reliable thermal monitor.
- Consider using a higher-quality cooling solution if possible.
Let me know if you’d like more details or further assistance!
I’m updating this thread because I’m still trying to understand what’s happening. I added screenshots from MSi Central—set PBO to Thermal Point 65—and am using EXPO profile 1 with User Scenario set to Silent. Not sure if those settings are correct or if other factors like voltages are the issue. My RAM is Corsair Dominator Titanium 6000Mhz, with 30CL and 2x32GB. It’s currently on the QVL for this board. If I switch the User Scenario to Balanced, most settings stay the same except the frequency fluctuates between 3800 and 5000Mhz. Changing PBO to Auto, Enabled, or Disabled causes temperatures to rise above 65°C, reaching 67-70°C at idle with spikes up to 85°C, even hitting 91°C during boot. This seems unusual. I also ran a clean Windows install, but it didn’t help. It’s just a guess—maybe there are BIOS settings I need to adjust, or my AIO stopped working after changing the MOBO.
Verify your CPU speeds both in the BIOS and operating system. I’m aware some boards offer an auto overclock option which I dislike—they often push your clock even when idle. I prefer to set a fixed speed and let it run, as it can get excessively hot without that adjustment. The older board likely lacked this feature, and I’m certain it was enabled by default—it worked for me.
Sorry for the loaded post - I'm just trying many things at once to solve this issue on my own Thank god someone finally replied. Do you know how this feature could be named? I also was suspicious that something like this could be the culprit, but I'm not so sure anymore. Here's what I did during my wait for some help: 1. I ditched the MSi Central settings 2. Switched my power plan to Balanced 3. Reinstalled chipset drivers for AMDs website 4. Set PBO to auto 5. Set CPPC to Driver 6. Found some info about AMD Cool'n'Quiet, I've found that in my Bios something similar is named Global C-state Control, it was set to auto, which supposedly functions as enabled, but I also switched it to enabled None of the above changed the temps 7. Checked the processes and CPU usage, nothing weird pops up, basically all processes are below 1% usage and the CPU in idle is around 2-6%, with 2-4 cores working 8. Checked the temps with HWInfo - CCD1 and CCD2 are both at ~52C average. CPU (Tctl/Tdie) is at ~62C average with a max of 70.4C Core temps are at ~44C for CCD1 and ~48C for CCD2, with max temp being around 4-5C more, apart from one core - Core7 on CCD1 with max of 55.3C. CPU Hotspot at 58C with max of 59.5C, no weird readouts 9. This is the weirdest one - I've set all my fans to basically 100% speed, NOTHING, literally zero change in temps, still 60 degrees and I've run this test for a solid 20 minutes 10. Lastly checked my AiO - there is no significant difference between the tubes, one is warmer just enough to know which is which, no weird sounds even at 3000RPM pump speed I'm attaching images of what HWInfo reads in terms of clocks, usually it shows all cores at 5Ghz, but sometimes 4-5, one or most cores go down to 3,3Ghz. Ryzen Myster however, shows that idle clocks are between 300-800Mhz. At this point I'm almost convinced that this is either some weird Bios setting like you said, or my AiO suddenly decided to give up on life. EDIT: Quick edit in regards to number 9 - it almost seems like something similar to PBO Thermal Limit is enabled by default. EDIT2: Did some Cinebench R23 runs: First run with both push and pull fans at ~1050RPM and pump at ~2800RPM - Multi core score was 29409 (23 passes), single core score was 1820 (2 passes). Multi core instantly hit and stayed at 89C, clocks for the first 15 passes stayed between 4100-4200Mhz, then 4000-4100, single core never exceeded 5000Mhz between C13 and C14, but kept jumping in the range of 3000 to 5000. Second run was at full cooling capacity, push fans at ~1800RPM, pull fans at ~1450RPM, pump at ~3000RPM - Basically the same as before, 29735 multi core score and 1821 single core score at same amounts of passes. Also instant 89C in multi core with no more than 4200Mhz across all cores and also no more than 5000 Mhz between C13 and C14 with same frequency jumps. Is the cooler causing thermal throttling or what is happening here? EDIT3: I decided to give it one last chance and changed the thermal paste and remounted the AiO again - now temps are 7 degrees higher, 67C idle... I decided to grab ID-Cooling FROZN A720 since it's on sale for just 55USD, should arrive by friday. If it's not the coolers fault then I have absolutely no idea what the hell is going on, maybe this PC is possesed for all I know. Edited July 8, 2025 by Chaoson Additional test and info for possible solution
This device is really pushing performance to its limit, isn't it? It's known as core performance boost, and you'll likely find the BIOS handles it well. Search for a YouTube tutorial on disabling it and manually adjusting your CPU clock for your specific motherboard. Edit: I found a helpful video that shows how to proceed. Using BIOS and AMD software for overclocking (plus undervolting) along with monitoring hardware temperatures and stress tests will help you stay safe. If temps remain high after stabilizing clock speeds, a slight undervolt can bring them down noticeably. Good luck!
Things shifted - I successfully managed the clocks, as shown in the images, and the CPU appears to function properly now. It actually runs faster than before. Still, the issue isn't resolved; it's getting worse. Even when the clocks aren’t stuck at 5Ghz, the CPU stays around 60°C during idle time. Moreover, during game loading, the CPU hit 89°C and my PC shut down right away. Is this thermal throttling caused by the cooler failing, or is there another factor at play?
I'm not completely confident about using Ryzen Master, so I'm struggling here. If I were you, I'd just remove the program and rely on BIOS for controlling clock speeds. I understand there are overclocking options and if I had to guess, it's about changing the motherboard settings. You can check your temperatures and speeds with hwmonitor. Your average clocks are lower, but they still run high under stress. There might be another adjustment needed. If you keep facing problems, I can share my BIOS configuration.
I reinstalled Windows once more, but this time I omitted the MSi Center, which likely didn’t interfere with the system. I’m fairly sure it wasn’t responsible for the current problems. Nowhere does it seem to be the root cause. My PC continues to sit idle between 60-65°C and reaches 89°C during startup, stress tests, and game loading screens. Despite hitting its thermal threshold, it still cools down. Interestingly, after leaving the machine off for nearly 15 hours, the first boot returned to 89°C but gradually dropped to around 51°C. The longer the machine stayed on, the higher the temperature rose, eventually settling in the typical 61-65°C range without ever dropping below that. I adjusted my Curve Optimizer settings, and the issue persisted at -20 across all cores. The system remains stable; pushing it further seems unnecessary unless long-term cooling is still an issue. Core timers appear to function correctly—CPU reaches 5.2GHz during games, Cinebench multi-core briefly spikes to 4.6GHz before settling near 4GHz at 89°C. Gaming sessions lasted two hours in Rematch, maintaining 80°C throughout and reaching 87-89°C during load screens. I used Ryzen Master for monitoring; it seems more reliable than HWInfo for real-time data. After contacting Corsair, they suspect the AI-O failed and suggested a replacement cooler. I’m still awaiting delivery of the new cooler to confirm further improvements. If neither Windows nor the cooler resolves the problem, I’m inclined to believe the BIOS settings are the main culprit (to be clear, I’ve followed optimized defaults, changed CPPC, Curve Optimizer, and thermal limits, but haven’t altered RAM). I’d appreciate a look at your BIOS configuration if possible. I plan to disable Core Performance Boost and report back after testing it again. EDIT: With Core Performance Boost off, the CPU never exceeded 89°C during boot (reached 65°C), which is a positive outcome. Cinebench multicore still hit 89°C gradually this time, taking a few seconds to stabilize. Idle temperature remains steady at 60°C. Locking core clocks at 4.2GHz is typical when Core Performance Boost is disabled, since that’s the maximum observed. EDIT2: I removed HWInfo and Ryzen Master, using HWMonitor instead—HWM behaves unpredictably. It shows normal readings while my system reports otherwise. EDIT3: I re-enabled Core Performance Boost, adjusted PBO Limits (defaulting to Disabled), which lowered the threshold. Still, idle temperature stays around 61°C. These processors tend to overheat and boost aggressively when needed. Synthetic tests like Cinebench can instantly hit thermal limits, yet real-world spikes still cause discomfort. I’m sharing screenshots from my BIOS for reference. Also, note that during idle cooling, the coolant temperature is about 41.7°C, rising to 51°C after two minutes of Cinebench—this is significantly above normal and may point to cavitation or air pockets inside the system.
You're following the correct approach. It seems the issue might relate to the pump header connection, but your pump speeds indicate the aio is functioning properly. Even if you push the system hard, the cooler should handle it well, which supports Corsair's recommendation. They've consistently provided reliable solutions, like the new kit when my RAM was unstable. Your BIOS shows two auto-setting options—Precision Boost Overdrive and Core Performance Boost. Turning them off lets you manually adjust the clock speed beyond your 4.2 GHz cap. Since your CPU model differs, your speeds are likely lower (around 4.0 GHz), but most of these settings can be transferred. Precision Boost works well at idle but may cause overheating under load due to excessive clocking. I can share screenshots once I return. Regarding the hardware monitor, it probably stems from Corsair software; HWiNFO is an option if you'd like to test it. If you have an AMD GPU, AMD tools let you track temps and performance. Task Manager shows clock speeds and usage but not temperatures. A slight underclocking on your CPU can still help—AMD CPUs gain from this. I lowered my CPU temps by about 12°C.
I also dealt with excessive analysis of the problem and kept searching every potential "what if" instead of verifying my AIO setup initially. You're learning, and I picked up quite a bit. I ran some extra tests with Core Performance Boost disabled, concentrating solely on AIO behavior. Last night I played Valorant—the lightest game I've had installed—went to the practice range, checked coolant temperature on my other screen, which dropped from 35.4°C to 43°C in just three minutes, while CPU temperature rose from 63°C to 73°C without slowing down. I opened the PC case: the area near the tubes was hot, the opposite side was cool, and both tubes felt warm with a strange hissing/grinding noise from the pump. After leaving the PC for about 13 hours, I started with a coolant temp of 25.4°C, which reached 35.5°C in 8–10 minutes. CPU temperature climbed from 49°C to 56°C, though it's worth mentioning I'm feeling quite cold (14°C) and the PC was near an open window. Cinebench multi-core caused the coolant to jump back up to 52.1°C within five minutes again. There was also a Bios temperature above 70°C constantly. Once I replace the cooler, things should stabilize, so I'll turn Core Performance Boost back on, keep PBO in Advanced with an undervolt of -20 or higher, and we'll see what happens. Still curious about your BIOS settings—HWInfo works well for me and matches Ryzen Master readings, so no issues detected. Almost certainly it was the AIO that failed. Thanks, ZROCK! I'd still be chasing software fixes if it weren't for you. The PC would probably be fine either way.