Device powers down during startup or when mid-game. Uncertain about the issue :(
Device powers down during startup or when mid-game. Uncertain about the issue :(
Initial details: Windows 10 64bit (fully updated) CPU: Ryzen 9 5900x (default settings) Cooler: Corsair h150i 360mm AIO Mobo: Asus Crosshair VIII Hero AM4 (BIOS 4601, recently updated to assist with my problem) *Virtualization and DOCP activated) RAM: Corsair Vengeance RGB Pro 32gb (2 x 16gb) DDR-3600 CL18 GPU: Aus Strix Gaming OC RTX 3090 24GB (default settings) PSU: Asus ROG Thor 850p (850w) Storage: 1x WD Black 1TB PCIe 3.0 x4 NVMe M.2 SSD (OS) 1x Kingston v300 120gb SSD (used for virtual machines during university) 1x OCZ Trion 460GB SSD (Steam) 1x Samsung 850 Pro 512GB SSD (old OS drive I retain in case I missed any updates) 1x Samsung 850 Evo 1TB SSD (games and video files from when I used YouTube) 1x Seagate Barracuda 2tb 7200 RPM HDD (general storage for unnecessary data) Case: Corsair Graphite 780T Also possessing 6x Corsair iCUE QL120mm fans (3 front inlets & 3 on bottom radiator), 1x Corsair iCUE QL140mm fan in the exhaust plus 3x Thermaltake Riing 40.6 CFM 120mm fans on the back of the AIO radiator for additional airflow. Additional notes: I still have an Elgato HD60 Pro connected if I ever need to run old consoles and a 4x port USB 3.0 expansion card. Also, I own 3x 27-inch monitors (2x 60Hz 1080p BenQ & 1x Asus 1440p 144Hz). Running software during the problem: Google Chrome, Discord, Steam, Game (Crashes in CS2, Apex Legends, PUBG as my main games; no crashes in Minecraft), Spotify plus background effects like iCUE, GeForce, Razer – not Cortex, and Asus Armory Crate. For the past week or so, the issue started when friends wanted to replay PUBG. The game loaded normally but would abruptly shut down when I clicked an in-game button. After reinstalling, it worked until mid-game it would repeat the same failure. No crashes during BSOD, no entries in event logs. Recently, it began affecting CS2 and Apex Legends as well; I haven’t experienced problems with PUBG recently. I’ve tested the CPU on CINEBENCH and found it normal. Installed FURmark for GPU testing, ran briefly (15 mins) without shutdowns, and checked hardware info during gameplay. So far, it seems like a power supply concern – the PSU should handle the load, but a previous BSOD hinted at memory issues that resolved themselves. The GPU reaches temperatures near 75°C in CS2, while hardware reports a hotspot around 105°C. Though uncertain, I suspect the problem lies with power delivery or cooling. About two weeks ago, playing Apex and CS2 without issues, but since then PUBG has become increasingly unstable. I cleared all PUBG files during reinstallation per an online guide. Notes: My motherboard has 8-pin and an extra 4-pin connector for the CPU, but I’m not sure this affects things. The system performed well months ago with two monitors and no issues. I’ve attached some error codes from Event Viewer that appear only when the PC shuts down mid-game, except during normal boot – no warnings there. Occasionally, when the PC powers off, I must switch off the PSU at the back and wait until it’s completely off before restarting. This also happens when I hibernate or perform a full shutdown. Any advice would be greatly appreciated. Thank you!
I experienced the same problem with identical motherboard and setup but uncertain if we share the root cause. In my case, the issue arose when using four RAM sticks versus two. I resolved it by adjusting the DRAM Current Capability setting in BIOS from the default 100% to 120%. Running Memtest86 showed no errors, though prolonged testing might trigger the problem. Now heavy overclocking is safe for RAM. For me, tools like Cinebench and Blender work fine. However, when launching PUBG or Apex, the system would shut down (since 2019 it didn’t).
A complete chaos! I also wager on Ram, turn off DOCP temporarily, leave only what you must.
Edited October 20, 2023 by leclod
This setup has worked smoothly since I activated DOCP, no issues encountered previously.
In my situation it worked even without DOCP enabled, give it a shot.
Adjusting the DRAM frequency to 120% hasn't caused any crashes yet while loading and playing CS2. However, I'm still worried about the temperatures shown in the hardware info—those numbers seem to match what's happening inside the game. I didn’t snap a photo while the game was running, but the GPU temperature was around 70°C and the hotspot was in the 100s. I plan to try PUBG and other titles tomorrow.