The dad's computer shuts down during gameplay in Sea of Thieves.
The dad's computer shuts down during gameplay in Sea of Thieves.
Hello everyone, I've been working on this for a few days now. My dad's setup includes an I7 8700k processor, 16GB RAM at 3000Mhz, a Tforce Vulcan Z, an Asus Tuf Z370 Plus, a Rog Strix RTX 3080, and an 800w power supply (80 Plus Bronze). Recently, he can only play Sea of Thieves for about 30-60 minutes before his PC restarts completely without any warnings or errors. It only happens with this game, while he plays Borderlands 3 and Tiny Tinas Wonderlands without issues. I've stress-tested all components with OCCT, Furmark, TestMem5, and everything seems fine. The graphics card drivers were updated too, and his temperatures are stable. I'm not sure what else to check—maybe looking at his BIOS? Thanks for any advice!
I suspect something is activating the safeguards on the PSU. I had a 1080Ti with a voltage module that only worked in Ark to make it respond exactly as needed. On the other hand, it could be that the PSU is deteriorating slightly, making it more reactive rather than completely failing.
It seems the restart might not trigger the protection circuit in the PSU as expected. Normally, it would shut down completely and require manually turning the power back on. It could also indicate a faulty, low-cost PSU. What could be causing this?
They had to admit what was actually occurring. I’m not grasping much of it, but summarizing what Gigabyte mentioned about the voltage module and EVGA’s comments on the PSU. I sent a return shipping label for the PSU first, thinking it was the issue, then replaced the GPU after the new PSU worked.
The suggested PSU should be around 850W. GA102 GPUs, featuring the RTX 3080, are known for significant power surges that could trigger overcurrent protection in the PSU if it lacks adequate capacity. High-end RX 6000 models exhibit similar behavior, often reaching 200% TDP and causing overcurrent protection trips. In the meantime, you can test this by configuring the GPU's TDP/power limit to 75% via software such as MSI Afterburner. If you stop experiencing random restarts, it strongly indicates a power supply issue, likely tied to the specified specifications.
Undervolting could also be a viable option Edited August 9, 2023 by leclod
I was considering this same problem—800w seemed sufficient, but I was worried about the power supply. It’s a generic Chinese brand, so maybe they’ll just swap it out. Luckily, he joined me at LXT due to seeing the Silverstone booth and was surprised, so I’ll test that power limit tonight to check if it helps. He was attempting to reduce graphics settings to see if it would use less from the card.
The only certain method I understand is reducing its maximum power limit. If the card features sharp spikes above 150% TDP, capping it at 75% should stabilize it to around 113% TDP during those spikes, assuming the proportion holds. An 800W quality power supply works well, and only RTX 3000 and RX 6000 models have shown this problem before. Prior to that, you could determine the ideal minimum power supply and it would suffice. So far, the RTX 4000 series hasn’t had this issue, with extensive testing by Igor's Lab confirming power spikes in the higher RTX 3000 and RX 6000 models. The RTX 4090 doesn’t seem affected, as confirmed by their research. I haven’t encountered any problems with the upper RX 7000 series yet, so I’m not investigating it further. The guideline is to overspec by about 1% or use a very high-quality power supply at the lowest spec.