BIOS update = worse stability?

Hardware Hub Motherboards
,
1

Hello! I did something yesterday that I probably shouldn’t have done seeing the results I’m getting. I decided to work on my system and one of the steps I went back and forth on was updating the BIOS, both for safety and some improvements Asus was touting in the release page.

So I decided to do so. In preparation I reset the BIOS back to stock, unplugged everything but a basic keyboard, the USB drive needed, monitor and ethernet cable. The update went through just fine and the system rebooted. After it rebooted I started punching back in all the previous settings I was using. First the fans, no issues. Than the RAM, no issues. Than the undervolt, no issues. All of this apparently because after I was using the system for a while I remembered that I had to turn the USBs off when the system is off and AURA was incorrectly set for my preference. So I went into the BIOS, turned USB standby power off and rebooted.

After I did that all hell broke lose. The system was struggling HARD to post. It was turning on and off a bunch of times and didn’t post. After 5 minutes of tries it finally posted and the Windows repair screen appeared. I skipped it and got into Windows just fine. I tried an OCCT CPU stress test (medium dataset, AVX2) and the system is stable. So I decided to put the CPU voltage back to stock and now the system posts more reliably, but still needs one try from time to time, especially if it’s booting cold (power unplugged, system at room temperature).

In changing all the settings I noticed something that struck me as weird after the fact: when applying the DOCP profile the list of changes looked to be much much shorter than it used to. If I’m not mistaken the motherboard was changing some other voltages and notifying me of that (maybe SoC, maybe other ones, can’t recally exactly). With this new BIOS it didn’t. So I started wondering if that might be the issue as to why the system is struggling to boot with a CPU undervolt that’s stable in Windows and maybe even posts when the chip is physically hot.

What could be the issue? Have I done something to cause the issues? How would you go about making the system more stable?

Specs: Asus Crosshair VIII Impact, BIOS 4601, Ryzen 7 3700x, 2x16 G.Skill F4-3600C16D-32GTZNC (on the QVL, if that matters at all)

P.S. I pulled up ZenTimings and looks like all the settings that the DOCP profile previously applied were applied just the same. Even VSoC has been slighlty bumped to 1.09 somethings (it fluctuates).

P.S.2. I tried to wake my system from sleep but it doesn’t even stay on for enough time to reset the BIOS. This is getting weirder and weirder…

2

One factor I know is that the undervolt settings might change (especially CO) with updates. CO -5 on agesa XYZ is not necessarily the same as on another agesa version.

How does the system behave if you revert the USB standby setting? It is strange that this has such an impact on stability…

2 Likes
3

I didn’t know that it could have such impact on the CPU. I was running a flat -0.075V offset and I would’ve never imagine that an AGESA update would make me lose such a low negative offset. I should try with a less aggressive offset, after I solve the cold boot issues.

I wasn’t indirectly implying that the USB standy power changes made the system unstable. I was just giving a time frame to the issues that I’m starting to have. I need to revert that change and see what happens. If it solves the issue I’m gonna lose it!

1 Like
4

@quilt I found the culprit. For some reason the DOCP profile doesn’t train correctly anymore. Without the DOCP profile applied the system behaves exactly like it did before the BIOS update.

@wendell Sorry for the tag, but do you know anyone at Asus I could talk to? I don’t mean to bother you or your contacts, but this BIOS update really messed up my system and calling support didn’t help at all. I know what’s wrong with it and did all the tests I could. Thanks!

2 Likes
5

I went a little further with my troubleshooting: after swapping the two sticks around the system can boot easly into Windows. But after putting it to sleep and resuming a couple times the issue still appeared. But this time the system boots to Windows and posts fine, than it power cycles.

1 Like
6

Ouch, Are any of the bios updates incremental? I just know i did a few jumping straight to current and it really messed things up. I’m sure you know this already, but worth the mention.

I know for me, i used to have to manually set voltages to get memory timings to work and not have random shut downs.

With the newest bios as of mid aug this year I think and I no longer had to do this. I still have the slower boot for training each boot, but I can live with it now that the systems stable with all stock voltage settings, docp II (Uses all setting from memory programed oc, dosent use asus optimised settings) 64gb 2 dimms at 6000, and using -15 and -25 for diffrent cores on a 7900x3d.

I do however have the asus x670e strix board which im sure is diffrent.

1 Like
7

Nope, I carefully red through all the BIOS versions and there was no in between version required to jump to this version. So I went straight for the latest one.

So is it possible that power cycling is triggered by DOCP even if the system has booted up already?

I think that was happening with the old BIOS I was running. Fans all spinning to 100% and than boot. Now, at JEDEC at least, it boots incredibly quick.

I really don’t know what to think, how is that even possible!? I think this is a big screw up from Asus really.

Well yeah, different architectures (mine is a Crosshair VIII Impact X570) but I think they should be as capable as it comes for the respective platform. Especially for memory!

1 Like
8

So, in my case yes… but it was more than a power cycle… for me it was like the pc went to sleep and wouldnt wake up. It was weird with the accessories on…but only some like leds not the fans.

I had to manually cycle power at the psu.

Have you tired a reflash with a redownloaded image? Maybe something got corrupted?

9

pull the battery out rather than hitting the reset button/jumper.

turn the pc on and off at the power brick, and short the battery terminals with residual power.
it takes about a second.
put the battery back in turn on the pc.
your bios should be set to default.

when you go into bios, load failsafe or optimised defaults. save and reboot.
back into bios…
set it up as you normally would. minus any oc profiles on the cpu.
save and reboot… see if your ram train’s properly at the extended profile.

good luck.

1 Like
10

It’s a laptop style battery double sided taped under the I/O shield, can’t really get to it unless I rip apart the board

Does it really makes a difference taking out the battery instead of using the clear CMOS through the button?

Thanks! Hope to find some at some point hahaha

11

I did not and, at this point, I’m really afraid to touch the BIOS and brick the board once and for all.

1 Like
12

ah i thought it a desktop…
jumper or reset button it is then. :frowning:

2 Likes
13

It a desktop :sweat_smile: But Asus decided to pack this board so tight that it didn’t have a space for a battery holder

Immagine
Immagine1920×1920 396 KB

The battery is under the circled area in this picture

2 Likes