There is a constant issue with inconsistent reboots during games, please assist.
There is a constant issue with inconsistent reboots during games, please assist.
Running the Ryzen 5 3600 with the stock cooler, ASUS TUF Gaming x570+ (current BIOS version 2607), 16GB g.skill Ripjaws v series at 3600MHz but BIOS set to 2133, PowerColor RedDevil RX 5700 XT (drivers 20.8.2), and EVGA Supernova 650 G+ (80+ Gold), along with a 3 Corsair AF120 in suction mode: DC mode Chasis fan1-3 (silent, standard, turbo - top to bottom), and 2 Corsair AF140 on CPU_OPT (DC mode, 55% under 30C, 75% under 65C, 99% at 75C). I also have a Samsung 970 Evo 1 TB M.2-2280 NVME SSD. Windows 10 is fully updated as of today, and I recently performed a clean reinstall, updating chipset, GPU, audio, and network drivers. I consistently experience crashes in COD MW, only rebooting with a black screen before login. The game runs smoothly at minimal to high settings; GPU temperatures stay below 76°C, with utilization ranging from 30% to 99% depending on conditions. Reboots occur almost every time I play. I’ve faced reboots during Witcher 3 and Apex, but setting all parameters to low still caused complete restarts. I’ve tried adjusting settings like changing Windows configurations, enabling D3D11 in extra command lines, running as administrator, and modifying memory scaling to .5, yet nothing resolves the issue. I’ve never encountered a crash or reboot outside of gaming sessions, suggesting Windows is operating correctly.
If anyone has suggestions, I’d appreciate hearing them.
Install OCCT, run a stress test while observing the time/voltage curve, and take a screenshot. Also conduct a Memtest to check RAM health.
Memtest performed well, no RAM issues detected. Cinebench and Unigine 4.0 also ran smoothly at maximum settings without crashes. OCCT is enabled and ready for testing, though I have some questions about the plot configuration. It automatically chose certain nodes but didn't select many others. Should I stick with its auto selections or verify all nodes are chosen? The screenshot taken before any tests was captured.
The thread discusses whether setting the voltage mode to Adaptive causes the voltage to fluctuate according to CPU load without exceeding the vCore limit. It also mentions understanding Offset mode, where the vCore acts as the maximum load voltage. The conversation explores BIOS settings for adaptive voltage and checks if deactivating them restores normal behavior.
link provided contains three brief evaluations (under 10 minutes). occt _Large data set w/o AVX, LINPACK, OCCT small data set w/AVX. Attempted GPU test with shader complex; reached 100% usage and crashed right away. Regarding the adaptive voltage, after reviewing more consistent outcomes in the link, do I still have to check this setting? OCCT CPU tests
Updated: I attempted to rerun the GPU test at 80% and again, experienced a 2-4 second black screen followed by a system reboot. The OCCT MEMTEST screenshot was used. Completed 12 iterations with approximately 278,000 errors detected. The POWER TEST also resulted in about 1 minute before another reboot. This outcome is really disheartening! The first build might be around three months old.
Do you happen to know what fan the AUXFANIN4 graph corresponds to?
Can you open the Windows Event Viewer (see relevant article on howtogeek) and check for any entries around the crash times?
I don’t recognize which fan is off by default. I’ll identify it and let you know.
The three 120mm fans are connected to mobo cha_fan1 (top front), cha_fan2 (middle front), and cha_fan3 (bottom front). The two 140mm fans are plugged into cpu_opt, cpu fan, and i have the stock coolermaster fan in aio_pump. All are operating in DC modes (they aren’t PWM fans). The front fan settings from top to bottom are currently silent, standard, and turbo. Both 140mm fans use the standard profile.
Windows event view shows numerous errors. Some occurred around previous crashes and others since I restarted everything. It’s difficult to pinpoint the cause. After turning the machine back on and running GPU/MEM/POWER tests, there were 2 winlogon_information events, 17 windows error reporting_information events, 17 ESENT_warning events, and 2 application error_ERROR events in WINDOWS LOGS>application.
Additionally, I attempted to recreate the reboot during GPU, MEM, and POWER tests, running each for 2-4 minutes without shutdown or errors. I’ll review the logs during the next crash I experience.