I've put together a temporary PC using an LGA2011 setup with a no-name Chinese motherboard (shows up as OEM X79G), a Xeon E5-2650v2, 64GB DDR3 ECC RAM, and a Vega 56 graphics card. Everything runs great during individual stress tests (CPU, GPU, RAM, VRAM). However, when I play modern games like Bloodlines 2, the screen goes black, the GPU fans stop spinning, and the audio cuts out, but the system remains powered on.
I've noticed that lowering the GPU power limit increases system stability. If I undervolt and downclock to keep the GPU at around 90-95W TDP, it works much better. I swapped out my power supply from an FSP SPI PRO 500 to an EVGA N1 750W, but it didn't improve the situation. I'm concerned about the motherboard showing bizarre voltage readings—like 12V at 6.67V and 3.3V at 5V—yet everything runs, which seems off. Could my building's poor electrical system be to blame? Other appliances like kettles and microwaves work fine, so I'm confused. Also, temps on the CPU (~50-55°C under load), GPU (72°C and hotspots at 85-87°C), and VRM (~60°C) are acceptable, though I'm a bit worried about the RAM running at 58-60°C.
2 Answers
It sounds like there could be an issue with the GPU itself, especially since undervolting seems to help with those crashes. Have you ever reapplied thermal paste on it? The VRM might be overheating, which GPUs typically don’t report. I'm betting that's worth checking out!
I built my first workstation using an X79 LGA 2011 back in 2012, and by 2019, it was struggling. The issue often isn’t about raw performance—it's more about using older components and the lack of driver support. Also, those Chinese boards might not be reliable because they’re often made from recycled parts. Just something to consider if things keep going south!
Performance is mostly fine for what I play, but I really want to make sure nothing is failing. If any part is toast, I’ll replace it, but it'd help to pinpoint what's actually dead.
I haven't reapplied it myself, but the seller said they did. I’m considering doing that again just to be safe, especially if the VRM is running hot!