2.A60 for B850 Tomahawk Max Wifi V1 doesn't like Reboots out of windows 25h2

[mgfirewater]

I do a lot with ram oc. and my setting was always fine. i stayed for nearly one year on 2A.40.
now after alle the new vulnarys where fixed in may, i decided to go to the next non beta Uefi fw.

yesterday 2A.60 was release by msi.
i set all my old values for pbo and ram oc. the benchmarks are identical. karhu runs without any issue for 30min. for a longer test i didn't had enough time yesterday,
because 2A.60 doesn't like reboots. which values i shoud have to change or is it any general issue of the agesa?

 

[E-Hobbit]

Out of curiosity since I have the exact same motherboard + Win version and also just flashed A60, what do you mean by "doesn't like Reboots out of windows 25h2"?
Since flashing I rebooted 2 or 3 times so far with no issues.

 

[mgfirewater]

But maybe other ram modules. Other oc setting or only xmp or expo. I will downgrade to my old fw A.40

 

[mgfirewater]

I found the error! The resistance CA, CK, CS "ODT B" values for 2A.40 and 2A.60 changed automatically via the "Auto" preset!

The automatically assigned resistance values for 2A.60 caused a crash upon reboot!

these are the manually set resistance values derived from the "Auto" settings for 2A.40, which I had determined about a year ago.
The AI wanted me to set ProcODT Pu and Pd to 43.6. However, a successful reboot is only possible when running asynchronously with values of 40 and 43.6 (which correspond to the old "Auto" values). It is astonishing that the new AGESA version altered the resistance values—which I had previously always left on "Auto"—in such a way, and that, conversely, only the *exact* resistance values from the old AGESA 1.2.03c "Auto" settings work correctly with AGESA 1.3.0.1b!

 

[Averter]

Have similar reboot issues with exact same RAM.
My issues is with using EXPO Profile1. After setting it in BIOS everything works on first boot (6000MT/s). After reboot there is long memory training and values will set back to JEDEC defaults (4800MT/s) and in BIOS EXPO Profile1 stays/shows "Enabled" (but thats just cosmetic bug).

EXPO Profile1 (6000MT/s) vs after reboot defaults

 

[mgfirewater]

Try to set all the expo set resistance manual in your UEFI profil. Maybe it will not changed them at ram training.

 

[E-Hobbit]

BIOS 7E62v2A61 is out.
"Optimized with AMD EXPO Technology: Featuring Ultra Low Latency." sounds like bug fixing

 

[Averter]

BIOS 7E62v2A61 is out.
"Optimized with AMD EXPO Technology: Featuring Ultra Low Latency." sounds like bug fixing

Thanks, after initial testing A61 fixes my EXPO issue. EXPO Profile1 ZenTimings A60 vs A61 are exactly the same.

 

[mgfirewater]

BIOS 7E62v2A61 is out.
"Optimized with AMD EXPO Technology: Featuring Ultra Low Latency." sounds like bug fixing

This release notes was also part of a60!

 

[mgfirewater]

@Averter would you share your zentiming screen with a61. So we all could compare your 2a60 and 2a61 expo 1 resistances

 

[E-Hobbit]

This release notes was also part of a60!

Well one time it says supported and the other one it says optimized (aka bug fixed).

 

[plutomate]

what is the exact model name of that ram? do you have a picture of the memory sticker?
by the way, just notice new BIOS release,

https://download.msi.com/bos_exe/mb/7E62v2A61.zip

with even newer AGESA code ....

 

[mgfirewater]

kingston fury beast 6000cl30 dr 1,4v. is this image is enough?

otherwise I could give you taiphoon burner screenshot.

@plutomate which information you see at the label, which is not shown at zentimings. would be nice, if you can explain what are you looking for.

 

[Averter]

@Averter would you share your zentiming screen with a61. So we all could compare your 2a60 and 2a61 expo 1 resistances

As stated before - everything is exactly the same in EXPO Profile1 between versions, including resistances.

Spoiler: A61

 

[mgfirewater]

I compare 2A60 and 2A61 per smu checker and for granite cpus are nothing changed

 

[ericj159802e9]

Nothing has changed for the System Management Unit. Changes in other parts of the BIOS are not reflected in this tool.

 

[plutomate]

@plutomate which information you see at the label, which is not shown at zentimings. would be nice, if you can explain what are you looking for.

just wanna make sure it's the SK hynix die.
have you tried memtest with EXPO 1 settings only? Does it give error?

 

[mgfirewater]

i didn't get any errors with karhu (the better one of memtest+).
only the reboot was hanging with the auto values of 2.a60.

actual, i use me my manual values from 2.a40 under 2.a60 and everything is fine.
when i have time, i switch to to 2a.61 or the next 2a.6x .

but the msi uefi is much more disappointing to reload a profile of a previous uefi version!
gigabyte uefi updates was "much much more" easier to handle.
on my old b550 aorus pro, i could SIMPLE reload my uefi profil from the previous version from an usb stick.
the whole Fan profile was only two clicks, load and safe.
also boot order, deactivated igpu, pcie speeds and much, much more was imported directly.
and in 8 of 10 updates. pbo CO values, ram oc values, voltages were accepted!!

you finished he uefi update, import the fan profil, import the uefi settings, and you were ready for a test run, if everything is stable and performance is similar!
now i have first to set for every fan controller 6 values (4 temps, pwr and alert). Why...

 

[plutomate]

it's not new that you would likely see issue from time to time when trying to load profile from previous bios to the current one.
not sure if it's the same on other manufacturer, but it's usually due to BIOS layout change such as adding item, microcode update etc.

 

[mgfirewater]

But it should possible to ignore specific settings and give the user a log. Pay attention curve optimizer settings are incompatible with the new UEFI. But boot priority and fans should always possible.