AGESA 1.2.0.0 issues

[Daylight_Invade]

I am running an X570 I-Gaming which I have put in the 1.2.0.0 AGESA (BIOS version 3402).

Overall whilst things seem more stable in respect to things like USB and WHEA and NVME related issues, there are however still some strange oddities I have experienced. Keen to see what others running X570 or B550 are seeing.

In my case I have noticed that the board runs a lot hotter in idle. Running a stress test for example actually cools the processor with my AIO installed which is totally counter intuitive.

[DrSlump76]

I'm testing latest beta bios on my configuration and i have to report the same problems/behaviours seen on past bioses, starting from 2311:
Wheas and random reboots turning on DOCP or raising fclk above 1600Mhz.

My system gains stability if i set manually the following voltages:
CPU_SOC_VOLTAGE= 1,025V
DRAM_VOLTAGE= 1,35V
VDDG_CCD= 0,95V
VDDG_IOD= 0,95V
CLDO_VDDP= 0,85V
1,8V_PLL=1,81V

With above settings i can reach 1866Mhz Fclk:

FCLK=1866Mhz
MEM_LCK=3733Mhz
CPU_OFFSET_VOLTAGE = -0.050V

This the latest test done with CPUID:

[img" TARGET="_blank">https://valid.x86.fr/cache/banner/ix3viw-6.png[/img]"]

[schm0]

Are those random reboots happening at idle?
Did you change Power Supply Idle Control to Low Current Idle?
On my system this caused random reboots/shutdown with a kernel power error in event log.
I think the PSU must support this feature, some (older?) PSUs do shutdown the power when the current power draw is below a certain amount.

I'm also trying to run FCLK@1800Mhz.
For the most part it is quite stable but I get one WHEA error a day (mostly at idle)
What seemed to work on my system was to actually lower the SOC and VDDG but raise the VDDP voltage a bit.
Default SOC voltage on my system is 1.1V.
Working voltages are:
SOC: 1.05V
CCD: 1.00V
IOD: 1.00V
CLDO_VDDP: 0.950V

This also fixed some weird audio glitch I had. (Sound turns robotic like for a second)

[DrSlump76]

schm0 wrote:
Are those random reboots happening at idle?
Did you change Power Supply Idle Control to Low Current Idle?
On my system this caused random reboots/shutdown with a kernel power error in event log.
I think the PSU must support this feature, some (older?) PSUs do shutdown the power when the current power draw is below a certain amount.

I'm also trying to run FCLK@1800Mhz.
For the most part it is quite stable but I get one WHEA error a day (mostly at idle)
What seemed to work on my system was to actually lower the SOC and VDDG but raise the VDDP voltage a bit.
Default SOC voltage on my system is 1.1V.
Working voltages are:
SOC: 1.05V
CCD: 1.00V
IOD: 1.00V
CLDO_VDDP: 0.950V

This also fixed some weird audio glitch I had. (Sound turns robotic like for a second)

Hi, i tried that option but doesn't seem to help. Seems like fixing soc, ccd, iod and vddp voltages solves the whea errors problem.
My system worked well, with no issues enabling docp, until bios 2206. After bios 2206 it's impossible to use the system with fclk>1600Mhz.
With PBO enhancer disabled system performances are on par with 2206 bios, so i don't believe that tighten memory settings are getting my system unstable.
With pbo enhancer enabled my system flies up becouse of better sustained peak cpu frequencies.

[odeivogersm]

DrSlump76 wrote:
...Seems like fixing soc, ccd, iod and vddp voltages solves the whea errors problem.

This is my experience as well.

[DrSlump76]

odeivogersm wrote:
This is my experience as well.

Asus support says it's ram's fault. But my ram (not in qvl) worked well until bios version 2206.

[xeizo]

DrSlump76 wrote:
Asus support says it's ram's fault. But my ram (not in qvl) worked well until bios version 2206.

It's certainly not RAM by itself, but it can be FCLK behind these errors and it is coupled to the memory bus. As earlier has said, trimming SOC, PLL, SB, VDDP and VDDG may solve the issues.

I would say something happened with the AGESA on these newer V2 bioses, is it the new ultra fast boost introduced with newer builds of Windows in tandem with the new AGESA? WHEA often happens when a high load is ended. I wonder if these things happens under Linux as well.

USB issues looks like some bus is starved for volts.

[Silent_Scone]

Given these updates are currently tagged with a beta prefix, can only recommend those experiencing stability issues that weren't previously to revert back to an AGESA pre 1.1.9 and wait for official builds to hit the support pages. Majority of WHEA related issues will be linked to instability of the Infinity Fabric when pushing FCLK too far.

[DavyDee]

Silent Scone@ROG wrote:
Given these updates are currently tagged with a beta prefix, can only recommend those experiencing stability issues that weren't previously to revert back to an AGESA pre 1.1.9 and wait for official builds to hit the support pages. Majority of WHEA related issues will be linked to instability of the Infinity Fabric when pushing FCLK too far.

Exactly what I did on my C8I - back to BIOS v3003 and much more stable, cooler, and predictable outcomes - though the bench scores lag behind.
Can't wait for stable versions with the kinks removed.

[Rizlah]

i was getting whea errors, random reboots and massive vcore spikes (over 1.5v) while idle.

how i fixed mine was to set a vcore negative offset of 0.13 and enable curve optimizer, i am currently running all core -10. I have PBO set as well, but that doesnt seem to have any effect on these issues.

I am now running 3800/1900 on my memory cl16

r9 5900x
tforce xtreem 4x8gb 4000mhz
CH8W 3202 beta bios (1.2.0.0 agesa)

[zomgugoff]

C8HW BIOS 3204 update is out and apparently no-longer in beta.