cancel
Showing results for 
Search instead for 
Did you mean: 

X670 resource

Shamino
Moderator

ill use this thread to collect some new test bioses for the boards, maybe also to explain some less understood options

to disable cores ccd go here and choose ccd xx bit map down core.
each ones stand for an enabled core
best to disable from the back, ie:
110000
instead of 0011000
after selection press downcore apply changes or discard if made mistake

ocpak/octools

FAQ:
7950x not boosting pass 5.5G -> check that CStates is not disabled
Detailed Explanation on CState Boot Limiter


Test BIOSes:

new:
X3D OC Preset for those MB with asynch BCLK Support: (for simple slight perf boost for X3D)
97792

DOCP/EXPO Tweaked: (for simple timings tightening)
97793

strixe-e 1515 

strixe-f 1515 

strix e a 1515 

crosshair hero 1515 

crosshair gene 1515 

crosshair extreme 1515 

creator 670 1515

creator b650 1515

strix 650E I

strix 670 itx

 

 

for crosshair and strix e-e:

explanation of segment2 Loadline:

dualseg.jpg

customize a heterogenous loadline for a dual segment workload range.

example above shows loadline=L6 when current is in range of 0~40A, and Level4 when current is above 40A.

 

 

 


Adds for x3d

dynamic ccd priority switch with core flex, os / driver agnostic so win10 win11 ok

97403

97404

Algo as follows:
If condition reached and ccd0 specified, then check current mem/cache activity > threshold and hysteresis reached, if fulfilled then switch
If condition reached and ccd1 specified, then check current mem/cache activity <=threshold and hysteresis reached,, if fulfilled then switch
Default hysteresis =4

Can combine multiple algos for ccd priority so combinations are wide

works on non x3d too but of course senseless on it. detailed explanation here.

392 Views
2,100 REPLIES 2,100

MagicHp wrote:

I'm trying to find something common between all those issues, but even something as vague as that last idea isn't.
There has to be a link or even a common root issue to all these, right?


I'd want to know if anyone with issues is trying to run with their pre-0922 PBO offsets and voltages, 'cos they definitely weren't stable for me. I had to go back to stock/auto PBO settings, and find a stable point from there.

Also, need to re-enter all settings by hand. Don't restore from a saved .CMO file

Is there no eCLK on the proart board?

Can say that I was running a fairly stable build on 0805 for quite a while:

ROG Crosshair X670E Hero, Ryzen 9 7950X, Corsair Dominator Platinum 32GB @6000/CL36 (DOCP2). Full build here: https://pcpartpicker.com/b/kxcTwP

On 0805 I ended up disabling most AI-based overclocking features and had only the DOCP RAM profile enabled and Memory Context Restore enabled to reduce the boot time. Before that I was getting general instability with games/apps crashes and OS lockups. I couldn't figure out the root cause of this but after a lot of troubleshooting I put it down to the auto-overclocking features and just turned it off for now. As much as I would like to get the best out of my hardware with a reasonable overclock, I had to put my experience down to the typical early-adopter pains and assumed it would get better after some BIOS and Driver updates.

Since updating to 0922, it's been quite problematic. It was not stable on defaults and I couldn't boot into Windows initially. I had to go through a couple of CMOS resets to get anywhere. Also had trouble getting into the BIOS as it would hang with a black screen. I can pretty confidently say that MCR does not work - as soon as I enable it, I will 100% blue screen on login every time. I have to then clear CMOS, make sure I leave this feature off, disable all AI-overclocking features and then I will boot into Windows 10 64-bit no problem, albeit with the long RAM training on a cold boot. I basically have to neuter my setup for the sake of stability...

I have plenty of power and cooling on hand and I think my system is capable of much more performance than I am getting right now, but I don't think we are stable as yet as far as the BIOS is concerned and that's making any kind of automatic or manual overclock quite difficult for me.

Castillan wrote:
I'd want to know if anyone with issues is trying to run with their pre-0922 PBO offsets and voltages, 'cos they definitely weren't stable for me. I had to go back to stock/auto PBO settings, and find a stable point from there.

Also, need to re-enter all settings by hand. Don't restore from a saved .CMO file



In my case, when I tried 0922 I started from scratch, and didn't touch or activate PBO when the problems started. DOCP II was activated and Curve Optimizer set to -15 , nothing else.
It's after having encountered issues with these basic settings that I tried re-entering my previous ones. Didn't work, tried various different combination, always ending with either BIOS not booting (requiring a clear CMOS to access it again) with differents Q-codes (most not referenced in the manual) , a BSOD concerning ntoskrnl.exe during Windows boot, or BSODs with various reasons at the windows lockscreen.

The only time I managed to get into Windows long enough while using 0922 to do anything was with a fully default BIOS.
I did follow the same steps as the Bios Flash instructions posted earlier minus the Clear CMOS on step 4, since updating a Bios always meant starting with default settings on next boot to me ^^

I may have some time this weekend to try a few things if anyone has some idea to test, or maybe try the new 0923.

MagicHp wrote:
In my case, when I tried 0922 I started from scratch, and didn't touch or activate PBO when the problems started. DOCP II was activated and Curve Optimizer set to -15 , nothing else.

That, right there, the CO of -15, is going to be your issue (just IMO). Set that back to 0 and retry from there. Don't use Ryzen Master. In fact, boot into safe mode and uninstall it. The CO of -15 is the same as setting a PBO of -15 in the BIOS, except RM will apply it when Windows starts. My stable curve settings dropped from -15 to -5 with 0922.

So yeah, for those having troubles. Boot into safe mode, uninstall Ryzen Master, and try then and see if the system stabilises.

Castillan wrote:
That, right there, the CO of -15, is going to be your issue (just IMO). Set that back to 0 and retry from there. Don't use Ryzen Master. In fact, boot into safe mode and uninstall it. The CO of -15 is the same as setting a PBO of -15 in the BIOS, except RM will apply it when Windows starts. My stable curve settings dropped from -15 to -5 with 0922.

So yeah, for those having troubles. Boot into safe mode, uninstall Ryzen Master, and try then and see if the system stabilises.


If it was that simple, the issue would have been solved in minutes ^^
Of course the CO is the first thing I deactivated, and it didn't change a thing. That is why I mentionned trying many things in my post.

I might have been more precise, but I meant that I tried Just DOCP, then just CO, then just PBO, etc trying different values when my previous settings didn't work and after my first attempt at boot with 0922 which comprised of DOCP and CO -15. Once I managed to boot into Windows using Full default settings, I even tried to change just the Q-fan settings, thinking it might be a flash error or something that would lock up at any change, but nope, I booted fine with modified Q-fan settings. Changing just DOCP to 4800 and leaving everything as it is would prevent boot for me.
Hope that clear things up 😉

I tried Ryzen master once with 0805, it took 2 hours to set up an instable curve, removed it right away ^^
It wasn't installed when I tried 0922 and had those issues.

I updated to 0922 this morning, turned on PBO, medium boostsit, EXPO II. Havent messed with anything else. Before I updated i made sure to update windows then updated the bios and then got the latest chipset drivers from AMD, for some reason Asus doesnt have the latest chipset drivers last time I checked.

After doing all that and running cinebench for 30mins to stress test it and playing Hogwarts Legacy at max settings including ray tracing with FSR set to quality for a few hours I havent had a single issue so far.

Apart from the fact I seem to be unable to get higher than 5.2Ghz all core and single core only boosts to 5.6Ghz this was the same for me on 0805 and I assume this is to do with having 4 sticks of Ram.

Here is my full specs:

7950x
Asus Crosshair Hero X670E
4x16Gb CL36 6000Mhz Corsair Vengeance
7900XTX Red Devil Ltd edition
2Tb + 1Tb WD_BLACK SN850X
1500W Bequiet Dark Pro 12

Tbh I think making sure windows is updated and installing chipset drivers is a big thing

MagicHp wrote:
If it was that simple, the issue would have been solved in minutes ^^
Of course the CO is the first thing I deactivated, and it didn't change a thing. That is why I mentionned trying many things in my post.

I might have been more precise, but I meant that I tried Just DOCP, then just CO, then just PBO, etc trying different values when my previous settings didn't work and after my first attempt at boot with 0922 which comprised of DOCP and CO -15. Once I managed to boot into Windows using Full default settings, I even tried to change just the Q-fan settings, thinking it might be a flash error or something that would lock up at any change, but nope, I booted fine with modified Q-fan settings. Changing just DOCP to 4800 and leaving everything as it is would prevent boot for me.
Hope that clear things up 😉

I tried Ryzen master once with 0805, it took 2 hours to set up an instable curve, removed it right away ^^
It wasn't installed when I tried 0922 and had those issues.


Thanks for the detailed explanation. Something is definitely wrong, but not sure what. You should be stable using stock settings (ie. flash bios, and then clear CMOS, and then boot into Windows). The first load into Windows may BSOD until Windows sorts out what it's doing, but the second boot should work. It's interesting that you mention Q-Fan. I'd forgotten but I had issues with the manual Q-fan too. When setting Q-Fan manually it seemed to mess the BIOS settings up and I had to clear the CMOS to restore. Using the graphical based Q-fan interface was fine. I don't know which you used, but I do think that something isn't quite right with Q-Fan in 0922.

Also have you tried with the 5.02.19.2221 latest chipset drivers from here: https://www.amd.com/en/support/chipsets/amd-socket-am5/x670e

Castillan wrote:


Also have you tried with the 5.02.19.2221 latest chipset drivers from here: https://www.amd.com/en/support/chipsets/amd-socket-am5/x670e


To be honest, I'm not sure. I have those installed since their release, but I think I made my tests with 0922 before I updated, might be worth a check indeed.

About the Q-fan settings, I didn't use the graphical mode. But when i say I modified the settings, I only set it as PWM and Silent mode, not anything engineer-level either ^^
But yeah, if something is indeed wrong with Q-fan, I guess any change might trigger it.

I receive more fans and a few other (optional) components tomorrow to finish my build, I'll probably try again from scratch making sure everything is updated before starting.
I'll probably try 0923, or is that an even worse idea? 😉

Soupladel wrote:
Although it is something that is frequently mentioned, the method or process of applying a BIOS update is rarely at fault when problems occur, its usually either a physical fault with the hardware or incompatibility with the code that is at fault.


Nowadays there are indeed several security systems to ensure that a flash goes well or in the worst scenario there is usually a way to "direct flash" without access to Bios. That wasn't always the case and my heart still skips a beat each time I update a Bios out of habit ^^

I mentionned the possibility here because the issues might have fit that scenario and others did mention it. Let's cover every path we find just in case 😉

MagicHp wrote:

I'll probably try 0923, or is that an even worse idea? 😉


There's been a number of reports of things not running smoothly with 0923 here: https://www.overclock.net/threads/asus-rog-crosshair-x670e-overclocking-discussion-thread.1799959/pa...

I personally wouldn't update to 0923.