cancel
Showing results for 
Search instead for 
Did you mean: 

Crosshair VIII Dark Hero new, 3204 BIOS is useless

mpowerph
Level 7
Hi, I am just wondering, what did the company think, when they released this crappy 3204 BIOS? Absolutely awful... The memory system became unstable, I receive random shut downs, critical errors, WHEA, sometimes with 00 POST Code, when the only way to restart the board is to cut off the power supply. Sometimes takes ages to boot into Win, because there is no POST or there is an endless boot circle with various POST Code errors. There is no PBO activated, only PB and the XMP profile.

I have tried backflash one of the old, stable BIOS, but it does not work at all, it does not allow me to do it, I cannot use the old, stable one again. So, I am seeking advice, what to do, I want to get a rid off the crap. There was a good BETA BIOS in middle of Dec, it just worked fine.
6,884 Views
11 REPLIES 11

Silent_Scone
Super Moderator
Hello,

If you're getting stability issues with XMP then you might want to dial in the memory overclock. Memory instability from build to build normally indicates the overclock was conditional on the older BIOS too.

1. How were you testing memory stability on the previous build?

2. You need to state your memory kit.

3.WHEA errors also point to an unstable overclock.

4. You say flashback isn't work, in what manner? Are you renaming the cap file as instructed? Is the USB drive formatted to FAT32 and in the correct port?
13900KS / 8000 CAS36 / ROG APEX Z790 / ROG TUF RTX 4090

Thank you for the quick respond.

Right, so I have a BIOS flash back drive, I use it only for BIOS flash. It is FAT32 formatted, nothing has changed. BIOS file renamed, as stated. When I try the flashback, I receive a message. "Selected file not a proper BIOS!" I have saved all of the previous BIOSes for the mobo, so there cannot be problem with the BIOS files. I have tried to use the BIOS flashback button, it does no do anything, I am sure for the same reason.

I use TeamGroup T-Force Xtreem ARGB black, DDR4-3600, CL14 - 32 GB dual kit(B-Die ICs) and also a 5950X is installed into the mobo. For stability test, I have used Memtest86+ from bootable pendrive, also I tried Y-Cruncher and Prime95 for CPU stability test, and TM5 for memory test under Windows. I do encoding, around 6 hours per day, and there is no issue at all with DOC mod feature enabled.

By the way, there is no overclock in normal use, like games and idle state, only the memory XMP had been activated. I switched off the PBO, only the PB is enabled. I receive BSOD, random freezes, restarts, and black screen(00 or FF POST Code error) in idle status only. In games, everything is fine, under encoding, all good, all of the stability tests are running well for ages without error.

With BIOS version 3003 or the older 2601 everything was fine, I could push the memory beyond 4000MHz, if I wanted, but on 3800MHz with 1900 IF it was rock stable. There were two BETA BIOS also, those were stable too, but for memory OC, It needed more SOC and VDDG voltage.

It was stable before, also I had an MSI Tomahawk mobo until I received the Dark Hero, I use the same setup I used there. This madness started with the 3204 version BIOS. And I cannot understand, how it is possible, that under heavy load, there is no issue, after I leave the machine idle, and when I come back after half on hour, it has restarted, switched off, etc... On this morning, I could not make a successful boot for 15 mins. I needed to Clear the CMOS. After I loaded my saved profile, and everything was good again.
No it is working since 6AM under AVX intense encoding, it does fine. But I bet, when I finish, and I start surfing on the internet, or just leave the machine as it is in idle, I will get a random issue again.

It is possible, with the new AGESA microcode came off, that my CPU has hardware issue, as it happened the first line of AMD CPUs. I am also thinking to write to the AMD support, and I will ask RMA.

I had very similar experience with random reboots, especially on system idle


I've managed to flashback normally using usb stick. Just formated it FAT32, copied over renamed bios and held that button for few seconds
No other way to flashback a bios to older version, you can't do it from the bios itself. You need to use flashback USB port that is properly marked on your MBO backplate and then while the pc is turned off, you hold the flashback button that is also located on the mbo backplate for about 3 seconds and then the LED on the button will turn on and blink I think.. then just be patient. it takes few mins to finish

Also main thing is that you'll lose all your OC profiles this way that the bios will load defaults


I was also very dissapointed with the new bios and I even tried stock cpu settings, no pbo or anything, but didn't even bothered to turn of XMP profile

Escorpiom
Level 9
I'm in the same boat.
Crosshair hero VIII non-wifi with 4 x 16GB Samsung C-die, I can't boot at it's rated speed 3200Mhz.
I'm now at 3000Mhz memory speed but still it won't boot most of the time, F9 post code.

This board has serious issues, notably with 4 dimms high density memory installed.
Actually the first bad product I get from Asus, I've been using Asus boards since AMD K5.

Cheers.

Escorpiom wrote:
I'm in the same boat.
Crosshair hero VIII non-wifi with 4 x 16GB Samsung C-die, I can't boot at it's rated speed 3200Mhz.
I'm now at 3000Mhz memory speed but still it won't boot most of the time, F9 post code.

This board has serious issues, notably with 4 dimms high density memory installed.
Actually the first bad product I get from Asus, I've been using Asus boards since AMD K5.

Cheers.



F9 is a training code. Normally means FCLK instability. Either too high, or insufficient voltage (VSOC) for the given settings. The board failing POST stress tests is there to indicate there is an issue with the applied overclock.

Latest build with 3733 CAS14 with 4x8GB bdie here without issues. Boost behaviour will depend on the AGESA version.
13900KS / 8000 CAS36 / ROG APEX Z790 / ROG TUF RTX 4090

I had exactly the same issues with 3204 bios went back to 3202 which seems to be much more stable. I hope asus sees this so they can fix it.

Silent Scone@ROG wrote:
F9 is a training code. Normally means FCLK instability. Either too high, or insufficient voltage (VSOC) for the given settings. The board failing POST stress tests is there to indicate there is an issue with the applied overclock.

Latest build with 3733 CAS14 with 4x8GB bdie here without issues. Boost behaviour will depend on the AGESA version.


Yes I understand that.
Issue is I do not overclock at all. Memory is rated for 3200Mhz and corresponding fclock is 1600Mhz.
This does work just fine with two dimms. But as soon as I put in all 4 dimms, the system no longer boots and fails with F9 post code and finally post in safe mode.

Your 4x8GB B-die is probably not high density. Furthermore, B-die memory has been available for quite some time and accepts higher voltages.
My issues are with Samsung C-die and 4 modules 16GB each. High density.
This memory does not scale well over 1.35V vdimm.

But to test the suggested fclock instability, I did run the ram and fclock uncoupled. That is, ram at 3000Mhz and fclock at 1866 and it runs fine, so it really isn't an fclock issue. It should be stable at the rated speed, 3200Mhz + fclock at 1600. Soc voltage set at 1.1v.

There hasn't been a single bios that runs 4 x 16GB Samsun C-die at rated speeds.

Cheers.

I have a Dark Hero and haven't updated to this BIOS yet. However, I just want to chime in with others that even disabling PBO entirely the motherboard seems to still boost to PBO like frequencies. Without PBO I should never hit above 4.8ghz on any core (on a 5900x) based on how the normal precision boost algorithm is supposed to work. However with everything disabled I still boost to like 4.9 - 4.95 on some cores with voltages hitting like 1.46v briefly.

Turning PBO back on results in literally the exact same boosting results and cinebench scores that are effectively identical.

I do have the OC switcher on so perhaps that's messing with it. I haven't tried with it off yet. I'll do some experimentation at some point.

xeizo
Level 12
I have the CH8 WiFi, latest bios is stable for me but I have dialed in everything manual.

Things to look out for causing instability is FMAX should be disabled, too much negative Curve Optimizer causes idle reboots and too much Boost Override causes idle reboots too.

This has been stable for a couple of weeks for me

87905
R9 7950X - ASUS X670E Crosshair Extreme (bios 0803) - 32GB HyperX Fury Beast @6200c32 - Corsair H150i Pro - ASUS RTX3080 TUF OC V2 - Phanteks P400A - Corsair RM850 - WD SN850