Hey all, I was messing around with the render scaling setting inside of MSFS trying to get the best performance out of my Reverb G2 and 2080ti. I was trying to get a handle on how much render scaling affected the performance by selecting anywhere from 50%-200%. Sadly, when 200% was selected, the PC almost immediately froze and proceeded to hard reset. Upon reset, the system wouldn’t POST and continues to boot loop with the last boot code being a “VGA Code 94” before looping. I can’t even get into the BIOS.
I’m now in the process of replacing components to figure out what has been damaged. I replaced the 2080ti with an old 980ti and still have the same symptoms, no POST and “VGA Code 94”. So, I don’t think it’s the GPU. My next replacements will be PSU, MB, and/or CPU to try an isolate the failed component.
I’m not sure exactly what happened here, but wanted to warn others from increasing the render scaling this much. Obviously, I expected it to affect the system performance, but I never expected it to end in an expensive hardware failure.
If anyone has any ideas what may have happened, had a similar issue, or have some things I can try to remedy, please let me know. Thanks! I’ll try to update this thread with the outcome.
Quick update… I’ve tried everything I can without buying new hardware. CMOS cleared, BIOS updates, replacing my 2080ti with an older 980ti GPU, reseating all RAM and connections, etc. etc. Still the same symptoms as above.
Everything I can find online points to possible damage to the CPU (Ryzen 3950x) preventing proper PCI enumeration. So, I’ve ordered a replacement CPU as the first thing to try.
This system has been running for over a year. CPU cooler is still operating properly and temps have been fine. It is really concerning to me, both from an AMD and a MSFS perspective… that experimenting with a render scaling setting could have caused damage to a CPU. This shouldn’t be possible from both a software and hardware perspective.
I’ll update this thread if I can make any progress or identify root cause. However, if anyone is running similar hardware, please be careful maxing out the render scaling setting!
Mainboard is an Asus Crosshair VIII Formula… “VGA Code 94” and no POST is supposedly a PCI enumeration problem.
I replaced the CPU today with a new 5950x and she boots normally.
The old 3950x shows no visible signs of damage or overheating. The thermal compound was evenly applied and had no signs of degradation. The AIO CPU cooler was nice and tight and was still operating normally.
I’m not blaming MSFS at all for this failure. It was probably a CPU defect that was going to happen at some point when put under this kind of stress (this was an early 3950x I bought at launch a year ago). It was not overclocked or abused in any way.
I do find it interesting and crazy that a CPU can fail instantaneously under stress for no apparent reason. Perhaps there was a core temperature that spiked when I increased the render scaling to 200%? I’m not sure how that could happen with a cooler that was (and still is) functioning normally. I also would think thermal CPU protections would have prevented such a failure.
This is honestly one of the most mysterious hardware failures I’ve experienced and, certainly, the first CPU failure. Considering the CPU cooling was operating great for a year before this failure and is still functioning normally, I am going to have to chock this up to a defective early-batch CPU that was only pushed over it’s limit after being stressed by the settings I was experimenting with in MSFS.
I had an early 3950X purchased within a month of launch.
It failed in March, with random instability (blue screens in Windows and Machine Check Exceptions in Linux) AMD sent me a replacement which has been rock stable.
I suspect there was something up with the early 3950Xs.
No piece of software can do that if cooling is fine. As example: with CPU based video rendering the CPU runs hours in 100% load and there is no damage. Normaly also a CPU starts throttling if is becomes to hot.
You have possible warrenty, there is a chance that you got a faulty CPU.