F5F Stay Refreshed Hardware Desktop Device shutting down or freezing during idle mode.

Device shutting down or freezing during idle mode.

Device shutting down or freezing during idle mode.

P
PowerWower
Junior Member
46
06-06-2025, 06:54 AM
#1
My PC began behaving oddly after upgrading from Windows 10 to Windows 11. I encountered several lock-ups and freezes during the installation process, often freezing or shutting down before completion. Until recently, everything ran smoothly; I cleared the SSD before transferring it to a new owner and then attempted to install Windows 11. The same problems persisted, suggesting possible CPU or RAM instability. I ran MemTest and received clean results, so I adjusted the CPU voltage slightly—raising it to around 1.4 volts for better stability. This resolved the freezes and shutdowns, allowing a successful installation.

Afterward, I tried booting in auto mode, but the PC refused to start, emitting three short beeps and displaying a RAM LED during POST debug. After a CMOS reset, it booted normally with all voltages set to auto. However, when I manually set them to auto, it still wouldn’t start, though it accepted after a reset.

To verify stability, I performed an extended CPU+RAM test using default settings with OCCT. The process completed without errors and confirmed the system was stable. I then checked the BIOS settings, ensuring the CPU core voltage was set correctly at 1.425. After changing it again in BIOS, I ran OCCT once more.

However, a new issue emerged: as soon as OCCT began loading all cores, they dropped to 400 MHz. Despite the high temperature (35°C) and watercooling, there was no thermal throttling. No power limits were applied in either OS or BIOS. Task Manager showed base speed at 3.9 GHz, but actual performance was only 0.4 GHz. The CPU utilization graphs displayed only 10% usage, while OCCT claimed full utilization and 400 MHz per core. Ryzen Master confirmed all cores ran at 400 MHz with minimal usage (around 10%), and the power consumption stayed under 20W. BIOS reported core voltage at 1.425V, contradicting the BIOS setting of 1.425V—though the value was actually 0.7V in real time.

Additionally, before this incident, I had recently flashed a BIOS that was over a year old, prompting me to apply security updates. The hardware specs were as follows: AMD Ryzen 5600G, MSI MPG B550 Gaming Plus cooling, Corsair H100i 240mm AIO RAM (Corsair Vengeance LPX 2x8GB @ 3000MHz), and a Samsung 970 PRO 512GB SSD. The freezing episodes occurred regardless of XMP settings, reinforcing the need for CPU voltage tweaks.

Any tips or assistance would be greatly appreciated.
P
PowerWower
06-06-2025, 06:54 AM #1

My PC began behaving oddly after upgrading from Windows 10 to Windows 11. I encountered several lock-ups and freezes during the installation process, often freezing or shutting down before completion. Until recently, everything ran smoothly; I cleared the SSD before transferring it to a new owner and then attempted to install Windows 11. The same problems persisted, suggesting possible CPU or RAM instability. I ran MemTest and received clean results, so I adjusted the CPU voltage slightly—raising it to around 1.4 volts for better stability. This resolved the freezes and shutdowns, allowing a successful installation.

Afterward, I tried booting in auto mode, but the PC refused to start, emitting three short beeps and displaying a RAM LED during POST debug. After a CMOS reset, it booted normally with all voltages set to auto. However, when I manually set them to auto, it still wouldn’t start, though it accepted after a reset.

To verify stability, I performed an extended CPU+RAM test using default settings with OCCT. The process completed without errors and confirmed the system was stable. I then checked the BIOS settings, ensuring the CPU core voltage was set correctly at 1.425. After changing it again in BIOS, I ran OCCT once more.

However, a new issue emerged: as soon as OCCT began loading all cores, they dropped to 400 MHz. Despite the high temperature (35°C) and watercooling, there was no thermal throttling. No power limits were applied in either OS or BIOS. Task Manager showed base speed at 3.9 GHz, but actual performance was only 0.4 GHz. The CPU utilization graphs displayed only 10% usage, while OCCT claimed full utilization and 400 MHz per core. Ryzen Master confirmed all cores ran at 400 MHz with minimal usage (around 10%), and the power consumption stayed under 20W. BIOS reported core voltage at 1.425V, contradicting the BIOS setting of 1.425V—though the value was actually 0.7V in real time.

Additionally, before this incident, I had recently flashed a BIOS that was over a year old, prompting me to apply security updates. The hardware specs were as follows: AMD Ryzen 5600G, MSI MPG B550 Gaming Plus cooling, Corsair H100i 240mm AIO RAM (Corsair Vengeance LPX 2x8GB @ 3000MHz), and a Samsung 970 PRO 512GB SSD. The freezing episodes occurred regardless of XMP settings, reinforcing the need for CPU voltage tweaks.

Any tips or assistance would be greatly appreciated.

R
reactscarface
Member
156
06-06-2025, 06:54 AM
#2
Also consider AMD chipset drivers. Swapping RAM sticks in and out might resolve the issue.
R
reactscarface
06-06-2025, 06:54 AM #2

Also consider AMD chipset drivers. Swapping RAM sticks in and out might resolve the issue.

L
LarsMatena
Senior Member
602
06-06-2025, 06:54 AM
#3
All drivers were added immediately after installing Windows, which came before any testing was performed. This happened after the initial problems and before the 400 Mhz lockdown issue. At least one previous test had worked properly—running at full speed—not at the locked frequency. It seems unlikely the drivers are the main cause unless something changed within two hours. I installed the newest versions from AMD directly, using auto-detection, which I normally avoid. I usually install chipset and GPU drivers manually, but since they’re both AMD, I’m not sure. I’m not convinced drivers are the problem, so I’ll try reseating the RAM just in case.
L
LarsMatena
06-06-2025, 06:54 AM #3

All drivers were added immediately after installing Windows, which came before any testing was performed. This happened after the initial problems and before the 400 Mhz lockdown issue. At least one previous test had worked properly—running at full speed—not at the locked frequency. It seems unlikely the drivers are the main cause unless something changed within two hours. I installed the newest versions from AMD directly, using auto-detection, which I normally avoid. I usually install chipset and GPU drivers manually, but since they’re both AMD, I’m not sure. I’m not convinced drivers are the problem, so I’ll try reseating the RAM just in case.

K
Kikibug223
Member
144
06-06-2025, 06:54 AM
#4
I attempted to reinsert the RAM while also resetting the CMOS, but everything reset back to its original factory state. Only XMP and POST beep signals were enabled; I found debug LEDs helpful, but real beep codes were more informative. I made sure the 400 MHz lock-down was cleared before proceeding. For a quick check, I ran another memtest, which completed successfully with no errors. Back on OCCT, the system clocks settled into the expected range (4,2–4,4), running at around 60W with a temperature of about 45°C. Under load, voltage reached approximately 1.36V, and idle was 1.44V (auto mode). The one-hour combined CPU+RAM test passed under Auto instruction set, extreme threading, large data set, and variable load. However, about 15 minutes later the PC froze, restarted itself, and I observed the clock screen freeze for 40 minutes before shutting it down manually. I avoided directly modifying core voltage and suspect a BIOS issue caused the lock-down at 400 MHz. Instead, I switched to PBO, enabled advanced settings, applied a +2 offset in curve optimizer, and planned another OCCT test.
K
Kikibug223
06-06-2025, 06:54 AM #4

I attempted to reinsert the RAM while also resetting the CMOS, but everything reset back to its original factory state. Only XMP and POST beep signals were enabled; I found debug LEDs helpful, but real beep codes were more informative. I made sure the 400 MHz lock-down was cleared before proceeding. For a quick check, I ran another memtest, which completed successfully with no errors. Back on OCCT, the system clocks settled into the expected range (4,2–4,4), running at around 60W with a temperature of about 45°C. Under load, voltage reached approximately 1.36V, and idle was 1.44V (auto mode). The one-hour combined CPU+RAM test passed under Auto instruction set, extreme threading, large data set, and variable load. However, about 15 minutes later the PC froze, restarted itself, and I observed the clock screen freeze for 40 minutes before shutting it down manually. I avoided directly modifying core voltage and suspect a BIOS issue caused the lock-down at 400 MHz. Instead, I switched to PBO, enabled advanced settings, applied a +2 offset in curve optimizer, and planned another OCCT test.

Y
YoloGeek22
Member
152
06-06-2025, 06:54 AM
#5
Curve optimizer isn't functioning properly. It worked well during idle for six hours, but froze again afterward. After a restart, the desktop froze within five minutes before launching OCCT. Trying load line calibration to level 4 didn’t resolve the issue. I suspect the BIOS might be involved, though I’m less certain now. I flashed back to the previous version, but problems persisted—the system would freeze or shut down completely, only under load. Testing with OCCT produced no errors, and even core cycling didn’t help. As long as the load is low, the system either locks up, freezes, or crashes entirely. Event Viewer isn’t providing useful insights; it just notes that the PC didn’t shut down properly. My last thought was to disable XMP, set RAM to 2133, and see if things stabilize.
Y
YoloGeek22
06-06-2025, 06:54 AM #5

Curve optimizer isn't functioning properly. It worked well during idle for six hours, but froze again afterward. After a restart, the desktop froze within five minutes before launching OCCT. Trying load line calibration to level 4 didn’t resolve the issue. I suspect the BIOS might be involved, though I’m less certain now. I flashed back to the previous version, but problems persisted—the system would freeze or shut down completely, only under load. Testing with OCCT produced no errors, and even core cycling didn’t help. As long as the load is low, the system either locks up, freezes, or crashes entirely. Event Viewer isn’t providing useful insights; it just notes that the PC didn’t shut down properly. My last thought was to disable XMP, set RAM to 2133, and see if things stabilize.

A
AlexCG
Junior Member
41
06-06-2025, 06:54 AM
#6
Unfortunately, even with standard configurations and factory settings, the system continues to freeze or shut down when idle.
A
AlexCG
06-06-2025, 06:54 AM #6

Unfortunately, even with standard configurations and factory settings, the system continues to freeze or shut down when idle.