Question Asus mobos burning x3D CpuS?

[igor_kavinski]

Stupid question time!

Maybe some people have done that. But most don't do it for researching the ins and outs of overclocking on a certain platform. Most people just expect things to work after changing a few settings. Set it and forget it is how it usually works with Intel. And presumably, that's how it was with Zen 3 (I hadn't joined the forums when that was launched so I don't know what dramas transpired at that time).

 

[IEC]

Stupid question time!

What's stopping people from:

- taking note of all applicable voltages
- enabling EXPO, taking note of all RAM settings AND applicable voltages with it enabled
- disabling EXPO, manually introducing all the RAM settings with applicable voltages BEFORE enabling EXPO
- increasing the applicable voltages slightly until achieving stability

?

That's too much effort for the majority of users.

For those of us technically inclined we actively sought out memory using Hynix M-die or A-die ICs and used Buildzoid's timings as a starting point (there are others, but his is probably most widely distributed). His suggested "should be plenty and probably room to decrease voltage" setting for SOC was 1.25V

After seeing the DDR5 overclock spreadsheet from OCN I concluded even that vSOC was excessive and dialed back to 1.1V (1.11V actual). I note the actual applied voltage because in at least one motherboard's case in the ongoing investigation there was a difference in voltages that were set in BIOS and that were actually applied resulting in 1.45V+ applied...

 

[biostud]

Stupid question time!

What's stopping people from:

- taking note of all applicable voltages
- enabling EXPO, taking note of all RAM settings AND applicable voltages with it enabled
- disabling EXPO, manually introducing all the RAM settings with applicable voltages BEFORE enabling EXPO
- increasing the applicable voltages slightly until achieving stability

?

The problem is when you're not told that this is how your bios is working and that sometimes the bios did not show the correct voltages. Also, when you change one thing in your bios, then if the bios changes something else "automatically" it should tell you exactly what it changes.

Lets say I enable EXPO and that changes VSOC from 1.1V to 1.25V when on auto, then I should be told that will happen.

 

[H T C]

That's too much effort for the majority of users.

Fair enough.

It was supposed to be until a "proper fix" is found, anyway.

 

[coercitiv]

Stupid question time!

What's stopping people from:

TL;DR Uncertainty. You're not in control. The software does what it wants, and it has conflicting ideas on what it wants.

At least some motherboard makers have implemented code that automatically overrides manual settings. For example, on GB boards increasing DRAM voltage may or may not trigger an automatic increase in VSOC voltage. On top of that, various pieces of code may conflict with each other, as the newer UEFI versions were released with the intent of limiting VSOC to 1.3V max, yet some still manage to jump to 1.35-1.4V when manual adjustments are made to DRAM voltage.

Last but not least, you don't even know if the code path that is triggered when enabling EXPO will be completely neutralized when disabling EXPO. As some other forum members have said, the common practice to avoid mobo maker shenanigans on older XMP platforms was to RESET the BIOS to default, as some changes from enabling XMP were not even visible to the user.

 

[H T C]

For example, increasing DRAM voltage may or may not trigger an automatic increase in VSOC voltage.

I thought that could happen ONLY when VSOC was set on auto.

Last but not least, you don't even know if the code path that is triggered when enabling EXPO will be completely neutralized when disabling EXPO. As some other forum members have said, the common practice to avoid mobo maker shenanigans on older XMP platforms was to RESET the BIOS to default, as some changes from enabling XMP were not even visible to the user.

That's easy enough: introduce that step after "step 2" and before "step 3". Call it: "step 2.5" xD

 

[igor_kavinski]

I like how ASROCK BIOS shows what changes were made before rebooting. To the users of other branded mobos, who else is doing that?

 

[coercitiv]

I thought that could happen ONLY when VSOC was set on auto.

We all did... before this:

What a disaster of mythical proportions! Gigabyte BIOS, the latest version, is still a minefield:

• to manually set VSOC the user needs to enter the same value in two different menu locations
• when altering DRAM voltage to high values, VSOC erroneously goes to 1.35V and even 1.4V !?!?

It's the same behavior as the one I highlighted yesterday on the Z690 board from MSI. User input can potentially trigger automated changes by the BIOS that the user is not aware of. It's frustrating, and in this case, dangerous.

The video shows that not only is the code automatically changing voltages based on user input for other parameters, this behavior is also producing different outcomes based on the order or the state of other adjustments. So each state of the BIOS is unpredictable until mapped out by the user. It's like modding the elevator, which is supposed to be a simple finite state machine with easy to predict behavior, and making it a much more complex machine with hard to guess and hidden decisions. Fingers crossed it's still finite

 

[coercitiv]

I like how ASROCK BIOS shows what changes were made before rebooting. To the users of other branded mobos, who else is doing that?

Does it show the change in System Agent voltage when you enable XMP?

 

[igor_kavinski]

Does it show the change in System Agent voltage when you enable XMP?

Haven't tried the XMP kit yet. I had the EXPO kit running on it. It only shows what settings I changed.

 

[JoeRambo]

There are two problems in play in this EXPO/XMP and SoC voltage disaster:

1) The common case of "Auto" voltages changing in automatic (pfft) way when EXPO/XMP is enabled, and they shove insane SoC voltages to make sure to minimize motherboard returns due to RAM that is not working at XMP. There are tiny variations here where if EXPO asks for 1.25V of DRAM voltage they will set like 1.27V etc. Nothing new really/
Do note that nothing in "settings changed" will show that voltage was adjusted -> it was AUTO and continues to be AUTO after enabling EXPO or XMP or overclock or changing some other voltage. You are at mercy of MB vendor to apply sane voltage while achieving their obviuos goal to avoid RMA due to incompatible RAM.

2) The extra retard version of this problem in Gigabyte BIOS. They have disaster level of BIOS support for AM5, two places to enter same values etc. So when changing SoC voltage, it might not apply or apply not the value you expect it to apply. Again "settings changed" would not indicate anything sinister, you really need to check HwInfo64 to see how voltages are.
It's not limited to voltages really, there are other interactions with memory timings and variuos settings that are in two places and have unpredictable interplay between them. Except other, non voltage related places result in frustated user ( remember that there is NO functional safe boot in this year 2022 platform from AMD, so each failed memory training => battery reset ). So no big deal for AMD, only enthusiasts suffer and even there a lower rung of them will just apply Buildzoid's timings for their memory IC vendor and not push things anywhere.

 

[H433x0n]

Stupid question time!

What's stopping people from:

- taking note of all applicable voltages
- enabling EXPO, taking note of all RAM settings AND applicable voltages with it enabled
- disabling EXPO, manually introducing all the RAM settings with applicable voltages BEFORE enabling EXPO
- increasing the applicable voltages slightly until achieving stability

?

I would say something like 90% of people aren’t comfortable doing that. I’ve got nothing to quantify that number but I would think it’s not a stretch to say most people aren’t comfortable performing those steps on their system.

 

[mv2devnull]

I would say something like 90% of people aren’t comfortable doing that. I’ve got nothing to quantify that number but I would think it’s not a stretch to say most people aren’t comfortable performing those steps on their system.

Most likely. (I'm one of that presumed majority.) Choosing from XMP/EXPO from UEFI drop-down menu is "easy". Anything beyond that is "beyond comfy".

 

[GaiaHunter]

Haven't tried the XMP kit yet. I had the EXPO kit running on it. It only shows what settings I changed.

Gigabyte shows the summary of what you changed.

 

[CakeMonster]

New ASUS beta BIOS'es with AGESA 1.0.0.7a, ver 1412
1) Update AGESA to ComboAM5 PI1007a
2) Improve System Performance

Probably not recommended for people who don't want to risk being on the bleeding edge.

I loaded the very same settings that I used with 1410, and have around 0.5% performance degradation based on CPU-Z and PyPrime, but it might be within the margin of error. I have custom settings with regards to timings, voltage, and temperature anyway, so any behind the scenes voltage tweakings might work differently for different people.

 

[aigomorla]

Probably not recommended for people who don't want to risk being on the bleeding edge.

Your asking X3D owners this...
Why not just ask them to stop breathing.
I think it maybe easier for some.

 

[biostud]

Your asking X3D owners this...
Why not just ask them to stop breathing.
I think it maybe easier for some.

Maybe Asus and X3D owners, but most others should be fixed, and personally I'm not throwing a beta bios in my system when I have a stable tuned system where voltages are within spec.

 

[JoeRambo]

I am not updating anything either on 7950x's. Newer AGESA increases power consumption on idle very substantially, no point in beta testing any more BIOS'es to see what they have broke this time.

 

[Racan]

Just appeared

 

[Puffnstuff]

Notice how Steve called Asus a scumbag company.

 

[DAPUNISHER]

Notice how Steve called Asus a scumbag company.

I am going to quote myself

With the amount of press this is getting, there is a silver lining. The legendarily horrific Asus customer service won't be able to screw over customers without facing the wrath of the internet mob.

My money is on - Asus refuses to take responsibility/keeps passing the buck, and is screwing affected customers over.

 

[moinmoin]

And GN is going to dedicate a whole 3rd part to them! *rubbing hands*

 

[biostud]

I am going to quote myself


My money is on - Asus refuses to take responsibility/keeps passing the buck, and is screwing affected customers over.

They will "RoG" on as the bling-bling crowd follows.

 

[TheELF]

I am going to quote myself


My money is on - Asus refuses to take responsibility/keeps passing the buck, and is screwing affected customers over.

Let's see, if asus has any evidence that AMD never told them that the max safe Vsoc had changed before hand then asus might even sue AMD.

 

[DrMrLordX]

So back to Asus motherboards for a sec: has Asus released a non-beta UEFI for the x670e Hero since 1410 came out? I'm still waiting...