ROG CROSSHAIR VIII HERO(WI-FI) BIOS 2204 / Follow-up

[Barryboy]

MY system: 3900X
Crosshair Hero VIII
Asus 2080 ti
Corsair 32GB 3600MHZ memory
RMI 850 Power Supply
Lian LI 011 Dynamic
Fully water cooled (Corsair Hydro X)

Upgraded to bios 2204 and I have major issues. Random BSOD, unstable game play,
When I try to use DOCP for the rated speed my pc won't even boot, it shows 22 on the QLED
So for now the only way to get into windows is having the memory run at 2666 MHZ and leave everything on default on the bios....

Any advice would be welcome, but I'm having the worst experience since I installed my system last year August. There is definitely something wrong with this BIOS...

[KotOr368]

New version 2206 is UP . i have Impact and that has 2103 and 2206 so i wonder if 2204 was some kind of beta for Hero version

[squirrito]

Just noticed there is a new BIOS version for Crosshair VIII Wi-Fi published today, just minutes ago! Bios version is 2206, release notes are identical to BIOS version 2204 so unclear what is changed. Might dive in and test now...

Edit 1: Initial results not promising. Booted into 2206 with optimized defaults, but had a BSOD minutes afterwards without any heavy load. Minidump is non-specific, just says ntoskrnl.exe. Applied DOCP which for my 4x8GB CL14 3600Mhz Samsung b-die kit has a rated voltage of 1.45v. I noticed in both the BIOS and HWInfo that DRAM voltage is measured lower than that, between 1.44 and 1.48v. Have not crashed yet, but a 3dmark stress test failed which is not a good sign....

Edit 2: Bumped XMP voltage a bit from 1.45v (XMP) to 1.46v, and things seem a bit more stable? Have been gaming for a couple of hours without BSOD at least, although game crashed to desktop once. Not sure if related or just a Tarkov issue....

[Barryboy]

2206 was releases, installed and I was able to boot into Windows 10 right away with DOCP enabled. I will do further testing and see if I'm able to do my normal memory overclock to have better timings. But so far so good...at least for the first hour

[Sprayingmango]

2206 was releases, installed and I was able to boot into Windows 10 right away with DOCP enabled. I will do further testing and see if I'm able to do my normal memory overclock to have better timings. But so far so good...at least for the first hour

2206 is fantastic, new AGESA as well!! Thanks Asus for this random gift of a BIOS update.

[Silent Scone@ROG]

Just noticed there is a new BIOS version for Crosshair VIII Wi-Fi published today, just minutes ago! Bios version is 2206, release notes are identical to BIOS version 2204 so unclear what is changed. Might dive in and test now...

Edit 1: Initial results not promising. Booted into 2206 with optimized defaults, but had a BSOD minutes afterwards without any heavy load. Minidump is non-specific, just says ntoskrnl.exe. Applied DOCP which for my 4x8GB CL14 3600Mhz Samsung b-die kit has a rated voltage of 1.45v. I noticed in both the BIOS and HWInfo that DRAM voltage is measured lower than that, between 1.44 and 1.48v. Have not crashed yet, but a 3dmark stress test failed which is not a good sign....

Edit 2: Bumped XMP voltage a bit from 1.45v (XMP) to 1.46v, and things seem a bit more stable? Have been gaming for a couple of hours without BSOD at least, although game crashed to desktop once. Not sure if related or just a Tarkov issue....

Hello,

By looking too hard at VDIMM readings you are leading yourself up the garden path. If you need to apply more voltage for stability, than by all means apply it. However, the amount of vdroop you are seeing here will be low enough to be considered non-existent.

There will be enough noise to mask how little droop there is unless using the proper measurement tool with an adequate resolution, as the UEFI reading is being taken on the power plane. A physical measurement would be required from the socket to eliminate any impedance/power plan related errors.

It is more likely that a change in memory ruling between BIOS builds is causing instability. Increasing VDIMM will often drown that out, this isn't anything new or untoward.

[tommy7600]

Hello,

By looking too hard at VDIMM readings you are leading yourself up the garden path. If you need to apply more voltage for stability, than by all means apply it. However, the amount of vdroop you are seeing here will be low enough to be considered non-existent.

There will be enough noise to mask how little droop there is unless using the proper measurement tool with an adequate resolution, as the UEFI reading is being taken on the power plane. A physical measurement would be required from the socket to eliminate any impedance/power plane related errors.

It is more likely that a change in memory ruling between BIOS builds is causing instability, increasing VDIMM will often drown that out, this isn't anything new or untoward.

You said that so many times that this start to be boring But to honestly, I would agree with you except from BIOS'es 2xxx increasing voltage fixes the stability issue that didn't exist in e.g. 1302. So there must be something with DRAM voltage.

[tommy7600]

BIOS 2206 - the same issues that I have in 2xxx series.

[Silent Scone@ROG]

You said that so many times that this start to be boring But to honestly, I would agree with you except from BIOS'es 2xxx increasing voltage fixes the stability issue that didn't exist in e.g. 1302. So there must be something with DRAM voltage.

If you find it boring, feel free to present some evidence to the contrary as I mentioned to you in the other thread

See the latter part of my above post, increasing VDIMM is a remedy for drowning out all kinds of stability issues. If you wish to waste your time pursuing something such as vdroop on memory rather than dialing out the instability yourself, there isn't much more to say, Tommy.

Having had this board in my personal system since launch, I'm not having any difficulties stabalising anything I have done in previous builds. That isn't to say some memory rulings have not changed.

[tommy7600]

If you find it boring, feel free to present some evidence to the contrary as I mentioned to you in the other thread

See the latter part of my above post, increasing VDIMM is a remedy for drowning out all kinds of stability issues. If you wish to waste your time pursuing something such as vdroop on memory rather than dialing out the instability yourself, there isn't much more to say, Tommy.

Having had this board in my personal system since launch, I'm not having any difficulties stabalising anything I have done in previous builds. That isn't to say some memory rulings have not changed.

Currently I have stable system by simply setting the voltage to 1.36V. BUT I would like to simply use DOCP profile that sets the voltage to 1.35V as I could do in 1302 without issues.