Game freezes unexpectedly following purchase of 3070 units.
Game freezes unexpectedly following purchase of 3070 units.
Hello everyone, I'm facing occasional crashes without clear error reports. They occur regularly in Ready Or Not, Homeworld Remastered, Teardown, and GhostRunner. Only Ghost Runner provides helpful messages (DXGI DEVICE ERROR HUNG), and these problems began roughly a month after installing a used 3070. I ran DDU after the issue started, updated drivers, reinstalled Windows, formatted then reinstalled the OS again, but the problem persists. Also tried Memtest86+, everything appears normal. I'm really stuck and have reached out to EVGA about returning the card via RMA, though I'd prefer not to ship it. Any suggestions would be appreciated.
Specs: Ryzen 5600x Asus B350F Strix mobo Corsair DDR4 3000C15, 16 GB EVGHA 3070 XC3 Black Corsair 850M PSU
Edit: Corsair TX850M PSU
It seems you're looking at a used RTX 3070 graphics card that's only lightly used. Some folks modify the ROM to fit their needs, but I wouldn't consider buying it just because of that. You could verify it's not modified or updated, though slightly underperforming GPUs can still have issues. There are drivers available, but they might not work well with every game on your system. Running games in admin mode is an option, but it won't fix hardware problems. If you're stuck, consider returning the item through RMA.
The PSU you're asking about is likely the Corsair M series. It's a popular choice for gaming rigs, and some models in that line can handle higher power demands, including those with voltage fluctuations. Make sure to check the specific wattage rating and any listed transient stability features.
A better PSU for your Corsair RXMX 650 should work. It looks like you're not up to date on transient power spikes, but once you find a suitable one, remember to flag my answer. The GPU power draw is around 220W, so the RXM 650 fits well. Your 3070 uses about 270W and has no problems with the RXM 650 (which includes the USB sensor). Solution confirmed from the Evga website. Solution
I don't think the Corsair TX850M can manage the demands well enough. I'm not seeing any loss of video output or system OPP tripping; instead, games are crashing to Windows. This pattern doesn't match GN's testing results, suggesting it's probably not a power spike issue.
I'm referring to the component that doesn't completely power down or cause a crash—it feels unusual, yet it often resembles a full failure. A video crash is typically fixable. Over the past two years dealing with 3000 series issues, I've found it's usually the PSU. Even though it's a used card, replacing it might be the best solution, unless you have other options.
The issue appears to have cleared after testing some updated Corsair 3200CL16 sticks (2x16GB). Both my previous and new memory modules passed Memtest, so I’m not sure why it happened. The card remains eligible for RMA since it has less than a year left on the warranty, and I want to ensure all concerns are addressed. Also note that when my older brother built my PC, he installed the RAM in B1 and B2, not A2 and B2 as ASUS suggests. I don’t believe this is the cause, as I never experienced this problem with a 1600x or 1060. New memory is configured in A2 and B2. Perhaps if you swap their RAM with someone who has trouble, it’ll help confirm the problem.