Aorus x670E Master and Bios Hell!

[Bobins]

Was running B13d bios without problems. System ABSOLUTELY stable. Latest GCC seems to have fixed the persistent GCC crash when it was started.
GCC updated Bios to B13. Started getting random BSODs (different message every time). GCC then updated bios to B21.
NIGHTMARE!!!
Motherboard won't boot to BIOS then might boot but then when Windows 11 loads network not available. The next boot might get to Windows but disk drives other than boot drive not showing. 3 out of 4 boots show an AMD GPIO driver problem in Device Manager.
Reset Bios to Optimised defaults. Configured again.
Still no change. Motherboard randomly wont boot and has to have power completely removed before it will respond.
Most successful boots to Windows have some hardware non functional.
Managed to update bios to B22b.
System will now boot but still occasionally BSODs with random errors. (System is not overclocked).
The stuttering sound problem has returned again though as soon as GCC loads the performance library.
So I either don't have non-default fan control or fan control and stuttering sound.
For a premium motherboard...this bios stability is CR@P.
Want to go back to F13d bios but can't 

[shadowsports]

Sorry to hear of your issues.  I'm not one to upgrade if I'm stable.  I use default fan control in BIOS.  Love how quiet my system is.  The fans on my GPU rev up a little when I'm rendering photos or video, but that's about it.  I have a small radiator and low RPM case fans that are very quiet. 

I don't think you can downgrade without risk of bricking.  Its not worth it.   Open a ticket with eSupport.  Hopefully the problems will get addressed.     

[Bobins]

I didn't want to update my bios and GCC decided it was going to do it when GCC got updated.

After extensive testing I ended up in a situation where the motherboard would not even boot. This necessitated a complete CMOS reset.
I have discovered that the instability is related to memory and that if Fast Boot is selected and there is a memory issue, instead of the bios reporting the error, it hangs. The only way to get out of this situation is the CMOS reset.

My memory is Corsair Vengeance DDR5 7000. When I bought the memory, I knew I would not get a speed of 7000 using XMP as this was an AMD platform not Intel.
With bios v13d, a memory test would run 24hrs stable at 6800 (XMP) with memory voltage/timing settings on Auto.
With bios v13 the system seemed generally stable but with occasional BSODs.
With bios v21 the system was totally unstable and almost unbootable.
With bios v22b it boots but regularly BSODs. In all cases the BSOD error changed every time.

I have gone back to testing and discovered that the bios versions after 13d will not run my memory at 6800. Indeed, it won’t run at 6600 either. I’ve had to drop the speed to 6400 to get stability.
This means the bios versions after v13d have all reduce my system performance and caused system instability. ☹

Also, any setting of a manual fan profile in the bios or using GCC causes a buzzing noise and video stuttering at about 5 second intervals in playback of any video file. If the fan profile is set at default, this doesn’t happen.
If bios is set to a standard fan profile and GCC set to manual start, the buzzing doesn't occur but as soon as you start GCC and it loads the Performance file, buzzing problem occurs.

I need to use a manual fan profile to adjust my water pump speed by water temperature.

Gigabyte say I should be able to revert back to bios v13d but I'm nervous. Can't risk bricking the system. Has anyone here experience of reverting their bios?

[dmdilks]

My memory is Corsair Vengeance DDR5 7000. When I bought the memory, I knew I would not get a speed of 7000 using XMP as this was an AMD platform not Intel.

This means the bios versions after v13d have all reduce my system performance and caused system instability.

There is the problem memory. I have said this in past and saying it again. Faster memory doesn't really run any faster then the base speed of the memory. I buy the lowest speed memory if can with the lowest cache too. 

The only time you will really see any speed in memory is on a server. The normal person is not going see a difference between 4800 to 7000 memory. Just run your memory at base speed and see what happens. Turn off the XMP.

I have been dealing with memory for ever. My first memory was PC66. You know it might not be the Bios. It could be the memory. Plus on the GCC when I 1st install it before I update anything. I uncheck things like the Bios update. I only use it to install drivers. After that it is turn off. That way you don't run into any problems.

Any time I do any fan adjustment it is in the Bios. On fast boot I never use it. I really don't care if takes a few extra seconds to boot. So what I'm going to say start at the basic settings. You not going to see any system performance change.

One more thing the base speed is 4800. When you go in and change it to the XMP. You are now OC'ing the memory. What I would try is run 1 stick at a time to see if it is memory.     

[Bobins]

At 6400 I've run memtest86 for 24hrs and the memory reports no errors so the issues with the bios doesn't seem to be memory related. I've tried running it without XMP and I still have problems.
I've narrowed the audio/video stuttering specifically to the use of the default "auto" setting in the fan type when a manual fan profile is used.
If I use a manual fan profile (controlling my pump speed and setting the fan speeds dependant on water temperature) with the fan type set to "auto" i get the audio buzz and video stutter. If I set the fan type to the exact fan type (PWM or Vol) the the problem goes away.
This bios is unstable though. The motherboard doesn't always sleep or shutdown properly needing a complete power down to recover. I can tell when this state is present because the onboard LEDs stay illuminated when the system is off despite the bios setting in bios for LEDs Off when powered down.
Sometimes on boot the network isn't detected. This is only solved if I go into bios and use F10 to exit without making any changes.
Sometimes the system won't boot with a windows BSOD reporting ACPI BIOS Error 

Gigabyte say I can regress the bios back to an earlier version but I'm nervous about this because I've heard of others bricking their system trying it.
I'm beginning to wonder if the motherboard itself is suspect.

 

[dmdilks]

The one I don't understand how the GCC did the update on the Bios. I have install the GCC many times and the Bios update is always uncheck. The only thing you have left is to down grade the bios to what you had.

Plus if you are talking to support ask them if you do this will the board still be in warranty. If something happens with down grade. It looks like you can do the down grade. Because they do have it some times when you do a Bios upgrade can't go back. But this it looks like you can.

[CaTaPulT]

Hello.
I don't have a "Master" but a X670 Aorus Elite AX.
I have experimented with newer bios files and saw a slight decrease in performance, the board came with F9 bios, I tried F22b and F21. Wasn't impressed, so I rolled back to F9 and have been happy since.
I use the bios flashing option in the bios screen itself, personally, I don't trust GCC much.
I download the wanted bios file and unzip it to a USB flash drive, reboot, use the bios built in flash utility and you choose the bios file you want from the flash drive, it will then verify the integrity of the file, ask you if you want to proceed, it'll reboot and do another verify, then it will flash the new (old) bios to the board and then reboot.
Be patient during the reboot process, this can seem like the system has hung but give it a few minutes and the system will come to life once again with the newly flashed bios.
Once this is done, I normally go into the bios and choose "Load optimal settings" and save and restart. Then I go back into the bios again and set the things I need to set like EXPO, disabling onboard graphics, etc.
In the week I've had this new board, I must have changed bios'es 5 to 6 times (gigabyte support asked to for proof that F21 was slower than F9).
I use my system for VR only (Racing and Flight Sims) and every little bit of performance is needed in high res VR.
Some people are very worried about changing bios, now days it's pretty safe, only time it may be a problem is if the process is interrupted like a power outage. But then again, most boards these days have a flashback button on them (dual bios) so the board should be able to be recovered.
Take care
 
Regards: Jack

[dmdilks]

The thing is why people think you have update Bios & Drivers when you computers is running fine. More then 1/2 the topic's here are when people have updated the Bios or Drivers. Plus when everything was working fine.

The Bios & Drivers are there for people having problems. Not for people when they computer is working fine. People think because they are there it doesn't mean you have to update them.

Some people are very worried about changing bios, now days it's pretty safe, only time it may be a problem is if the process is interrupted like a power outage. But then again, most boards these days have a flashback button on them (dual bios) so the board should be able to be recovered.

This is from Gigabyte:

Warning:
Because BIOS flashing is potentially risky, if you do not encounter problems using the current version of BIOS, it is recommended that you not flash the BIOS. To flash the BIOS, do it with caution. Inadequate BIOS flashing may result in system malfunction.

On the flashback button I think it might be on the top of the line boards only.

I just work on a computer that was working fine for over 2 yrs. The guy is a I-T guy. He updated the Bios to put in a better CPU. Well from that point on he had nothing but problems with it. Plus he did it 3 times too. The last one was a Beta Bios and now you can't go back either.

The problem was the board even after the update saying it could run the better CPU. But the board still didn't like the CPU. I put in a different CPU and the board has been running fine. There nothing wrong the CPU it is running fine in another board.

The point I'm making here is:  "If it isn't broke don't fix it"