What is the PCIe Errors Recovery Counter?
What is the PCIe Errors Recovery Counter?
AER stands for "advanced error reporting". It is a capability found in certain PCI cards, where PCIe refers only to the physical layer—it remains part of PCI, but is transmitted in a serialized form. A PCI card can encounter various errors and either stop or fix them. I wouldn't dismiss this entirely; it depends on the specific error. The AER acts as a reference to a sequence of errors; a NULL value means no error detected. Typically, there shouldn't be any errors present. Possible reasons for these issues are many. For a more in-depth explanation (available under Linux), refer to:
AER within the Kernel
If signal quality is affected, the card may revert to a slower standard. If another issue exists—such as a mismatch between firmware and driver versions—these could trigger an AER. Possible causes include bit flips. The range of potential causes is vast, and you might not fully understand them without reviewing the error logs. You could try searching online for details on these errors:
how to examine pcie advanced error in windows
Unstable power, heat, or a poorly seated card are also potential factors. If I were analyzing this, I’d want to identify exactly which error is being reported and which PCIe slot it affects. It’s possible the issue isn’t with the card itself but with an embedded device.
This outcome seems likely. I kept encountering issues with the 5080 on the motherboard. Even after adjusting the BIOS settings—changing the iGPU "Auto" to "Forced" and then disabling it—I didn’t see any improvement. It’ll be fun to check if these problems disappear when I switch to an all-AMD configuration with the 9070XT Red Devil. 🤔
Only a single PCIe port is utilized. The Corsair RMe 1000W serves as a Tier A power supply, and the suggested PSU capacity is 850W according to ASUS guidelines. Heat isn't an issue since the card runs at about 30°C and GPU memory usage stays low around 40°C, with Florida's ambient temperature at 78°F. The issues observed stem from a cold start when the computer was powered down overnight.
It looks like an unusual upgrade from the 5080 to the 9070 XT, considering the current GPU prices. I own a Rzyen 9950X and 7900 XTX along with HWMONITOR, and it doesn’t show any comparable issues. This might not be related to your case at all—just a reference point.
Check out the Red Devil 9070XT versus the 5080 in Microsoft Flight Sim 2024. Also, take advantage of a great deal on the Red Devil for $849 from Amazon—compared to the MSRP of $1484.99 and ASUS's higher price of $1599.99, Amazon now matches the cost. You can return it either by card.
Did you remove the 5080 during testing or experimentation and also uninstall the Nvidia drivers as part of that process?