I think your PC shuts down unexpectedly because it lacks adequate heat management
I think your PC shuts down unexpectedly because it lacks adequate heat management
Hello everyone, I've owned my computer for two years and sometimes it restarts unexpectedly, especially during games. Occasionally it also shuts down while running normally. Online advice pointed to possible issues with RAM, overheating, or the power supply. Based on what I’ve tried, the RAM seems fine after testing it with Windows’ built-in tester. I think my PSU has enough capacity for the 3070 chipset it supports, so the main concern is likely cooling. Living in Puerto Rico means a hot climate, and running FurMark on the GPU pushed temperatures up to 92°C. My setup includes two 120mm front intake fans (AIO cooler for CPU) and two rear exhausts, plus one additional 120mm fan on top of the case. I’m considering upgrades to improve overall temps and hopefully fix the problem. Here’s what I have:
- CPU: Intel Core i5-12600K 3.7 GHz, 10 cores
- Cooler: Corsair H100x 57.2 CFM liquid
- Motherboard: MSI MAG B660 TOMAHAWK WIFI DDR4 ATX (LGA1700)
- RAM: G.Skill Ripjaws V 32 GB (2×16 GB) DDR4-3200 CL16
- Storage: Silicon Power P34A80 1 TB M.2-2280 PCIe 3.0 X4 NVMe
- Graphics: Gigabyte GAMING OC Rev 2.0 GeForce RTX 3070 LHR, 8 GB VRAM
- Case: Corsair 4000D Airflow ATX Mid Tower
- PSU: EVGA SuperNOVA 750 GA 750 W (80+ Gold, fully modular)
Any suggestions or recommendations would be greatly appreciated!
Furmark isn't meant for temperature testing. It's a synthetic maximum load on a GPU. Your cooling system works well per case. The AIO handles the 12600K too. It won't trigger shutdowns; it just reduces performance. The power supply might shut down if something tries to draw voltage beyond its limits. Possibly the GPU is failing, which could cause the PSU to stop. The 3070 might be deteriorating and force the PSU to cut power. Many factors can lead to this, including bad hardware on any part. Do you have another GPU or PSU available?
No, I don't have access to another GPU or PSU right now, but I guess I could ask a buddy to let me borrow theirs. But this comes with another dilemma. I'm not sure how to test the issue because when it restarts (which is weird why wouldn't it just shutdown?!) it's truly random it isn't always when a game starts or in the middle of one. But, why do you think it's a hardware issue rather than a temperature issue? Also, on the PCPartPicker list the estimated wattage is 494W even with a 750W PSU could the 3070 still spike and hit the PSU that hard? Edited August 25, 2024 by Thuer Added estimated wattage
I don’t have access to specific tools or software, but you can try setting up a controlled environment using common development platforms like Docker, virtual machines, or simulation frameworks. This will help you replicate and test the issue effectively with other components.
Not directly related to GPU problems. It's hard to pinpoint the exact cause. Unknown voltage settings, frequency combinations, etc. I encountered an issue with a 1080ti using a faulty voltage module. In Ark dud it triggers crashes, but the specifics remain unclear. This is just one example—other factors like RAM might be involved. Try testing one component at a time to check for memory issues?
What dimensions does your AIO have? Does it ever reinitialize when you're not engaging in heavy tasks, such as browsing or using a word processor? Consider this: take out your RAM, gently wipe the gold contacts with a dry sheet of paper, and use a fan to expel any debris from the slots. Inspect for any warped connections. Reinsert the RAM and test different settings that seem to cause resets. Try launching Prime 95 and Time Spyseveral times each, ensuring both run under high power and generate significant heat. If one triggers a reset while the other doesn’t, it may indicate a problem with the power supply or motherboard. For instance, if only the TS model resets, that suggests an issue with the GPU or power unit. Take out the graphics card, clean the gold contacts thoroughly, and store it aside. Verify the slot is pristine and free of bent pins. Connect your monitor to the motherboard and attempt a reset. If it fails, it’s probably the power supply, though the motherboard itself might be hidden. Also, use HWiNFO64 to track temperatures for any sudden spikes that could lead to resets. Let us know what you find.
The AIO features a 240mm radiator paired with two 120mm fans. This setup is uncommon, but it has happened occasionally. I’ll go through all your steps carefully and share the outcomes. Regarding step 3, "GC" likely stands for something specific—please clarify if you’d like me to explain further. Thank you for the advice!
GC refers to the graphics card. The GPU acts as the processor inside the card, but people often refer to it together, which can cause confusion.