X670E TOMAHAWK WIFI - BIOS 7E12v1F1 issues

m+ms154f02e7

New member
Joined
Aug 31, 2024
Messages
4
I tried the latest 7E12v1F1 beta BIOS to patch CVE-2023-31315, which is fixed in AGESA PI 1.2.0.1. It has a couple major issues:

1. Memory training only takes a couple seconds on DDR5 6200. Robust training is ON and MCR is OFF. This leads to a wildly unstable machine. I can get it to train properly for one boot by entering UEFI settings and saving, if I can get that far. No combination of MCR and DDR5 Nitro settings results in proper training behavior. This OC's rock solid on 7E12v1E.
2. An AQC107 NIC in PCI_E4 is either not detected or shows as being unable to start due to insufficient system resources in device manager. There's nothing in M2_4. My only SSDs are in M2_1 and M2_2. The only other PCIe device is my GPU in PCI_E1.

Anyone else seeing similar behavior?
 
What kind of performance improvements do you get when using Nitro settings?
Is it better than using DRAM High-Efficiency mode presets (found in MSI's OC settings)?
 
I had left the Nitro settings disabled on 7E12v1E. I only started touching the training related settings on 7E12v1F1 in an attempt to force the previous behavior.

I haven't tried MSI's high-efficiency mode presets- I manually dialed the timings in instead. I do have EXPO enabled as a baseline. The memory training's behaving as if I'm running a fully stock DDR5-5200 config.
 
Today's 7E12v1F release BIOS is still broken. The memory training's fixed but I can't use my PCIe NIC.

It's the same AQC107 NIC in PCI_E4. It works on 7E12v1E. In device manager, I get "This device cannot find enough free resources that it can use. (Code 12)."
 
I tried disabling the integrated WiFi, audio and NIC in the BIOS to no avail. Thankfully it works in PCI_E3. Not 100% optimal as it reduces GPU airflow there, but it gets me up and going. Also, TIL that slot doesn't share bandwidth with PCI_E1.

The AQC107 NIC is on the latest firmware 3.1.121 with driver 3.1.10.0.

ETA: Here's what's in my PCIe slots
PCI_E1: RTX 3080
PCI_E4 broken, now PCI_E3: AQC107 10GBASE-T NIC
M.2_1: MSI M480 2TB
M.2_2: WD Black SN750 1TB
No SATA devices
 
Last edited:
I have the same issue with AQC107 NIC, also on firmware 3.1.121 with driver 3.1.10.0 in PCI_E4, and it worked without issue on 7E12v1E. With 7E12v1G, The card works on every cold boot, but gives a "This device cannot find enough free resources that it can use. (Code 12)." after every warm boot, and once it happens, if you remove the device and rescan, it does not find a missing device, but rebooting it comes back with code 12.
 
It's now late March and I'm fighting the AQC107 NIC issue, but on my Carbon board. I will assume that it's the same issue and good to know it's device specific.

Disregard: I thought I had clicked on an X870e thread, my bad.
 
Last edited:
I updated to 7E12v1H, and the same issue still exists when the NIC is installed in PCI_E4. Works on a cold boot, error code 12 after a warm reboot. On reading advice, others have suggested moving the card to PCI_E3, and the results there are even worse, the card is never detected most of the time, and the one time I did, it fully disappeared (not even a code 12) after warm reboot.
 
Back
Top