Very Slow Bios Update. Should I restart?

[mclaren85]

Since for last couple of days I got the error code "94" I've decided to update my bios. But now, it is flashing very very slow as it has been roughly 2 hours since I started the procedure.
And I'm pretty sure why it is so slow. It's because of I use pci gen 3 compatible riser cable albeit my computer supports gen 4. And before the bios update, I clear the CMOS. The default values are set for gen4 so the image loads very slow than anything.
My motherboard's support link is down below, and I do need your help now. Should I wait for couple of hours, or just press the reset button?

https://www.msi.com/Motherboard/MEG-X570S-ACE-MAX/support

 

[VuurVOS]

Restarting the system while doing a bios update is always a bad idea since it will very likely brick your system.
Luckily you have a bios flash button on the rear I/O panel so it should be recoverable when you power off the system.

These are the instructions mentioned in the manual of your motherboard:

Updating BIOS with Flash BIOS Button
1. Please download the latest BIOS file that matches your motherboard model from the MSI® website.
2. Rename the BIOS file to MSI.ROM, and save it to the root of the USB storage device.
3. Connect the power supply to CPU_PWR1 and ATX_PWR1. (No need to install CPU and memory.)
4. Plug the USB storage device that contains the MSI.ROM file into the Flash BIOS Port on the rear I/O panel.
5. Press the Flash BIOS Button to flash BIOS, and the LED starts flashing.
6. The LED will be turned off when the process is completed

 

[GerKNG]

errors, broken bios flashing?
RMA it. it's too expensive to play around with.

 

[ir_cow]

BIOS flash takes around 5-10 minutes at the most. Something froze up. Only option now is to flash from the rear I/O port since the BIOS will be corrupted from the failed update.

 

[mclaren85]

After couple of hours (5-5,5) the update is completed without fault.
Yet I still don't know why the system had given me the 07 and 94 post codes errors prior the update.

 

[A Computer Guy]

Did you flash using USB stick? Perhaps something is wrong with your USB stick but you managed to get though it anyways.

 

[remixedcat]

the first stage on bios updates is always quick but the 2nd where it flashes the intel extras always tends to take longer than the manufacturer's suggested times... on this latitude 5400 the 1st phase was done in 1 minute and the intel part where it flashes all the management engine and vpro and other features took 5 minutes. same on my other laptops too that are intel based. My inspiron that has insyde h20 took 4:45 to flash.

 

[LabRat 891]

After couple of hours (5-5,5) the update is completed without fault.
Yet I still don't know why the system had given me the 07 and 94 post codes errors prior the update.

Did you flash using USB stick? Perhaps something is wrong with your USB stick but you managed to get though it anyways.

I may be incorrect but, I believe most boards will 'handshake' USB 1.x 1.5 Mbit/s 'low-speed' for BIOS flashing.
That limitation + an especially slow/troublesome USB stick, absolutely could be at (partial) fault.

For an example:
I have a 16GB SanDisk 'cheapie*' flash drive - it's about as slow as a floppy for file transfers. Never has thrown an error, just obscenely slow from Day1.
(*purchased Back-To-School sometime while Windows 7 could still be purchased on a new PC)

w/o looking up the post code errors:
I'm guessing you had either a corrupt or semi-compatible firmware on the board. (IIRC, moving to/from a UEFI/BIOS w/ Legacy Ryzen support can take 'extra long', too)
Perhaps, may have been part of the reason for the long flashing procedure.

(hopefully unrelated) I wonder if a failing flashROM would produce a similar 'issue'?

 

[ir_cow]

After couple of hours (5-5,5) the update is completed without fault.
Yet I still don't know why the system had given me the 07 and 94 post codes errors prior the update.

Praise the BIOS Gods! Time to play the lottery, your a lucky person.

 

[mclaren85]

The problem came out to be a faulty riser cable (gen4). Even after each vibration, the mainboard thinks there is a new hardware found and load optimized settings for BIOS. However as my riser cable only supports for gen 3; it freezes because of synchronization issue between gen3/gen4.
After removing the faulty part, it is working as it should.

Btw, I triple check my riser cable because I couldn't believe the cable is faulty since it is an ROG cable..

I hope why pci slot won't sag after using like this:

Praise the BIOS Gods! Time to play the lottery, your a lucky person.

I deserved it, cause I've waited with patience

*Update: After installing without riser cable, I just wanted be sure, so I re-update my BIOS. This time, it completed under just 2 minutes.

 

[LabRat 891]

The problem came out to be a faulty riser cable (gen4). Even after each vibration, the mainboard thinks there is a new hardware found and load optimized settings for BIOS. However as my riser cable only supports for gen 3; it freezes because of synchronization issue between gen3/gen4.
After removing the faulty part, it is working as it should.

Btw, I triple check my riser cable because I couldn't believe the cable is faulty since it is an ROG cable..

I hope why pci slot won't sag after using like this:

View attachment 342932


I deserved it, cause I've waited with patience

Wow! That's a new one (for me); that's going in my mental toolbox.

Also, 'goes along with' my experiences w/ 'that style' of riser. I've had both a 'well-built generic Gen3' and a Lian-Li Gen4 x16 riser "fail" after only a few uses.

If I ever have to riser-out/off an x8-x16 lane GPU (again), I'll be investing in Oculink or another Server/Industry SFF-# 'spec' interface.
[My MaxCloudOn risers, have been mostly-reliable but, they're Gen2-Gen3 spec. Oculink, etc. 'should' allow Gen4]

These 'high-frequency ribbon-like' risers, seem to get damaged solder joints from just looking at them wrong.

 

[mclaren85]

Wow! That's a new one (for me); that's going in my mental toolbox.

Also, 'goes along with' my experiences w/ 'that style' of riser. I've had both a 'well-built generic Gen3' and a Lian-Li Gen4 x16 riser "fail" after only a few uses.

If I ever have to riser-out/off an x8-x16 lane GPU (again), I'll be investing in Oculink or another Server/Industry SFF-# 'spec' interface.
[My MaxCloudOn risers, have been mostly-reliable but, they're Gen2-Gen3 spec. Oculink, etc. 'should' allow Gen4]

These 'high-frequency ribbon-like' risers, seem to get damaged solder joints from just looking at them wrong.

Just Relax I have some reasons why my riser failed on me. One of them was excess heat. My Riser cable was too long that I had to bend it. And it was being touching to nvme all the time which is constantly above 50 degree of celcius. The other reason was forcing the gpu back and forth while cleaning inside. That my well harm to soldered joints as you mentioned earlier on.

I don't think I will ever invest in riser cables again.

 

[Chrispy_]

Glad you sussed it.
I hate risers; I've used them for gaming builds, mining rigs, and compute servers and no matter how much you spend they all seem to suck!

I've had enough issues with Gen3 cables that I wouldn't consider a Gen4 build with risers any more. I'm sure the right combination can work but it just adds an additional layer of complexity and some of the case designs that call for riser cables also bend those cables with a horrible radius, which is just asking for trouble. At least you're not NZXT having to deal with a riser cable incident so bad that two separate governments had to step in and mandate a recall (https://www.pcgamesn.com/nzxt/h1-case-fire-hazard-recall)

 

[ThrashZone]

Hi,
Yeah my asus x99 used to toy with me a bit and show no activity or anything on the screen
I usually tongue in cheek and pushed the start button and it comes on telling me the usual, down power off blah....
Pretty annoying thankfully it's the only board ever to do that nonsense hehe

 

[dgianstefani]

With risers if you don't have one the same standard as your motherboard, just manually set the PCIe link speed in bios.

I run a gen 4 riser for my GPU since I have a sandwich case. Previously had a gen 3 riser. Zero issues with either.

 

[A Computer Guy]

Glad you sussed it.
I hate risers; I've used them for gaming builds, mining rigs, and compute servers and no matter how much you spend they all seem to suck!

I've had enough issues with Gen3 cables that I wouldn't consider a Gen4 build with risers any more. I'm sure the right combination can work but it just adds an additional layer of complexity and some of the case designs that call for riser cables also bend those cables with a horrible radius, which is just asking for trouble. At least you're not NZXT having to deal with a riser cable incident so bad that two separate governments had to step in and mandate a recall (https://www.pcgamesn.com/nzxt/h1-case-fire-hazard-recall)

Do PCB risers have the same problems as cable risers?

 

[c2DDragon]

I didn't read everything !

Should I restart during a BIOS update ?

NO

Go do your stuff. Don't do anything unless your motherboard is dual bios. :3

Ok now I will read everything

 

[LabRat 891]

Do PCB risers have the same problems as cable risers?

Considering their ubiquity in Industrial PCs, Servers, and Point of Sale machines...
No, But:

Some of Supermicro's risers are not PCIe standard pinout. (Thankfully didn't kill anything finding that out)
Some 'trustworthy' brands like Silverstone, *may* pull just a little too much current over the traces. (Power traces on my Raven's riser overheated and started pulling away from the PCB)
Some 'generic' PCB risers have horrific solder joint quality and/or are missing smoothing caps or, have poorly laid-out traces (limiting use to Gen1.x-2)

 

[Chrispy_]

Do PCB risers have the same problems as cable risers?

I've never had an issue with server PCB risers, which are usually just part of the motherboard and designed-in from the start. Servers with 7 GPUs in them still use cable risers though

 

[LabRat 891]

I've never had an issue with server PCB risers, which are usually just part of the motherboard and designed-in from the start. Servers with 7 GPUs in them still use cable risers though

Indeed.

Quality Cable Risers (as used in servers) are built very little like the ones commonly consumer-available, sold w/ cases, etc.




U.2, Mini-SAS, and Oculink PHY 'import' risers should be much more durable, as well.

 

[Calenhad]

The problem came out to be a faulty riser cable (gen4). Even after each vibration, the mainboard thinks there is a new hardware found and load optimized settings for BIOS. However as my riser cable only supports for gen 3; it freezes because of synchronization issue between gen3/gen4.
After removing the faulty part, it is working as it should.

Btw, I triple check my riser cable because I couldn't believe the cable is faulty since it is an ROG cable..

I hope why pci slot won't sag after using like this:

View attachment 342932


I deserved it, cause I've waited with patience

*Update: After installing without riser cable, I just wanted be sure, so I re-update my BIOS. This time, it completed under just 2 minutes.

If you attempt to run a gen3 riser cable at gen4 speed, you are literally trying to run it at twice the speed it is designed for. That is not a faulty cable, that is using it wrong. You have to reduce the max speed to gen3 in bios to have any chance for that to work properly.

Your Suprim card came with a GPU support in the box btw. You could use that to prevent sagging.

 

[Chrispy_]

Your Suprim card came with a GPU support in the box btw. You could use that to prevent sagging.

Those MSI ones that come in the box are useless if they're the ones that attach using the PCIe slot covers underneath the GPU.
I had those included with my 3070 and 3090 and it didn't do much for the 3070, the 3090 was too heavy and needed a foot down to the PSU shroud like OP is using.

 

[Calenhad]

Those MSI ones that come in the box are useless if they're the ones that attach using the PCIe slot covers underneath the GPU.
I had those included with my 3070 and 3090 and it didn't do much for the 3070, the 3090 was too heavy and needed a foot down to the PSU shroud like OP is using.

The Suprim cards have a free standing "pillar" support in the box. If you look at the product page you can see it in action. I would not say it wins any design contests, but it is a simplistic, working support that does not come with tons of RGB and what not.

 

[Chrispy_]

The Suprim cards have a free standing "pillar" support in the box. If you look at the product page you can see it in action. I would not say it wins any design contests, but it is a simplistic, working support that does not come with tons of RGB and what not.

Ah okay, they've learned from their mistakes - the 30 series cards I had were gaming x series and had these useless things in them:

 

[RJARRRPCGP]

just press the reset button?

No, don't! I suspect that's why there are stories on the internet of MSI motherboards getting bricked with BIOS updates easily! (people not waiting and assume it's crashed)

But in my case, for my MAG Z490 Tomahawk, it didn't take long like you describe.

I flashed a bunch of BIOSes this year. I was able to get the MAG Z490 Tomahawk to a 2023 version. All my ASRocks to a 2024 version (all socket AM4) and the same for my socket AM4 ROG.

The MSI MAG Z490 Tomahawk was the only one however, where there's no supposed "LogoFail" fix, MSI didn't mention that for the MAG Z490 Tomahawk, and it's from before 2024.