The system shuts down without obvious signs of overheating or software issues (resolved).
The system shuts down without obvious signs of overheating or software issues (resolved).
I believed it was just overheating from components wearing out after a couple of years. But: #1] The chipset has cooled down to about 55°C again after I added an extra fan. Its radiator is now warm. #2] I applied thermal paste to the CPU, which was almost dry, yet it didn’t bring the temperature back below 55°C. Its radiator remains warm. #3] The GPU’s radiator gets uncomfortably hot – but shutdowns now occur within minutes or so after cooling down overnight. Sometimes it works for a few hours and then powers off when I’m just browsing, while other times it stays on for 12 hours and shuts down after four such sessions while watching a 4K film at 75 FPS with SmoothVideo Project. After resetting the system, I couldn’t get back up because the hardware keeps shutting down before my desktop loads. Occasionally it runs smoothly for a few hours, but then it turns off abruptly. It’s not due to faulty software. I restored the system from an old image, disabled all Windows updates, and checked that everything was securely connected. I also swapped out the PSU socket used for both GPU and CPU. None of my drives draw power from the PSU, and I haven’t added any new hardware. So I suspect the PSU isn’t the issue – maybe a local workshop would help by testing it live? But what then? Repeat the same process for the GPU and CPU, leaving the motherboard behind.
Sorry, I missed the remaining parts of the setup. Here’s the full list:
System: Windows 10 Enterprise 20H2 x64 mobile processor: GIGABYTE X570 Aorus Ultra rev. 1.2 [BIOS F30] CPU: AMD Ryzen 7 5800X [100-000000063WOF] GPU: GIGABYTE Radeon RX 6700 XT GAMING OC 12GB [GV-R67XTGAMING OC-12GD] [BIOS F3] + riser Thermaltake Premium PCI-e 3.0 x16 Extender 1000 mm [AC-051-CO1OTN-C1UK000276] PSU: Seasonic Prime Fanless TX-700 700W [SSR-700TL] case: none [components on desk] drives:
● SSD NVMe M.2 2280 Samsung 980 PRO 500GB PCle 4.0 [MZ-V8P500BW / R-R-SEC-MZ-V8P1T0]
● SYSTEM ● SSD NVMe M.2 2280 Samsung 980 PRO 1TB PCle 4.0 [MZ-V8P1T0BW / R-R-SEC-MZ-V8P1T]
● SSD NVMe M.2 PNY XLR8 CS3040 4TB PCI-e Gen4 x4 [M280CS3040-4TB-RB]
● SSD 2.5" ADATA Ultimate SU800 2TB SATA III 560/520 MB/s TLC 3D [ASU800SS-2TT-C]
● SSD 2.5" Samsung 850 EVO 500GB SATA III 540/520 MB/s 512MB cache [MZ-75E500B/EU]
● SSD 2.5" Samsung 860 EVO 1TB SATA III 550/520 MB/s 1GB cache [MZ-76E1T0B]
● SSD 2"5 Samsung 870 QVO 8TB SATA III 560/530 MB/s [MZ-77Q8T0BBW / R-R-SEC-MZ-77Q8T0]
NVMe devices are linked through PCI-e ports, while SATA models use Unitek DiskGuard Limpid R USB-C To SATA Enclosure S1103D.
The machine restarts itself after a forced shutdown, possibly indicating overheating – it also powers down automatically after a night off. Recently, it resets without any BIOS changes for about two years. I haven’t attempted any overclocking to prevent this.
In short, the hardware and software have remained consistent for roughly a year, but this issue has appeared only recently and unpredictably – not following any clear pattern.
Not necessarily—I own a PCI-e card that can be used for an M.2 drive. But yes—I made an error because I thought my NVMe devices were connected via M.2 slots using my PCI-e bus lines. In reality, they might have been linked through an external SATA case, which I also possess. Regarding the installation process, I assume switching to a backup drive that mirrors its contents will keep things consistent. For testing without the full system, I’d probably just browse the web, hoping for a quick resolution. Even then, I’d likely restart immediately after a power cut or failure, as I’ve experienced before. Sometimes I can boot up fine after such an event, but other times it happens rapidly. As for the fanless PSU comment, it’s odd—I’ve been using one since around 2007 and it still has a solid 12-year warranty. A high-quality model from Seasonic in 2020 clearly isn’t worth the doubt.
Yes, I was firm. The issue likely originates somewhere and your configuration is unique. That power supply can overheat occasionally. You shouldn't rely on all drives every day. I also don't like your image-based explanation. It might be a repeat of an error. Could you skip the riser cable and use what you have? Make sure the core remains steady. Edited September 15, 2023 by leclod
I spent time, effort, and resources on purchasing a second motherboard with the same model and a different CPU. I also swapped out the RAM, but ultimately it was the power supply unit that caused these shutdowns. After installing a new PSU, the system functioned normally again.