Asus Z690 ProArt Creator WiFi (Thunderbolt 4) + i7-12700K + AMD RX 6800 XT

[fabrik]

What's the XMP like? Deffo not an extreme one
XMP 1 (one)

Which sticks?
Exactly match

...but you had two and then installed two more, right? So it's two sets, right ?
Ye, right, Correct. 4x16GB 6000MHZ

As per my testings, 2703 is a good one. 2802 might not be better, but it really depends on individual config. I'd say, get to 2703, but let me know which stick you have. Also, try this:

• flash 2703
• AI Overclock Tuner -> manual
• BCLK : DRAM -> 100:100
• Memory controller : DRAM -> 1:2
• DRAM frequency -> 5200 (we will work our way up from here: start at 5200, if it boots go to 5600 then, maybe, 6000)
• (scroll) CPU System Agent -> manual -> 1.2V (lower will do, XMP sets it higher, so 1.2V is absolutely okay for now)
• DRAM VDD / DRAM VDDQ -> 1.4V each
• (enter advanced memory voltage) IVR Transmitter -> 1.4
• Memory controller voltage -> 1.375
• PMIC -> sync all PMICs

Save and report back
So what I did, I flashed to 2703 and reboot it. It got the memory at 4200MHZ. So I turned the AI and left everything in auto. Rebooted it and Voilá = XMP 1 (one) with 6000MHZ.

Just trial and error. and reading lots. I get frustrated as well, no worries lol

Key takeaway: the point is not much enabling XMP or not, rather which profile XMP sets (and on which config). Because a XMP which sets speed at 5600MT is not really that heavy on the system. But the higher you go, the more difficult it is. Also, 32GBx2 is more intensive stress wise than 16GBx2. All being the same, 4 sticks are a hell lot more difficult than 2 sticks provided matched sets (possibly from the manufacturer, not “purchasing two identical sets”). Z690 is almost always worse than Z790. 13th gen is “always” better than 12th. 14th might not be better than 13th gen.
Then… there’s tweaking to mitigate the situation.

THank you!

 

[m4rk0]

View attachment 576105

THank you!

Wow that was better than manually tweaking it you had a decent degree of luck for booting 4 stick @ 6000MT on a Z690. But you should (at least) run OCCT in Windows with these settings: Large data set / Extreme. That's to be sure it is stable. Then a few hours of TM5 with absolut profile. The fact it boots does not mean it is stable.

Regardless, 2703 is a good BIOS because it made my 2x32GB run better at 6400+

 

[racermaster]

Thank you so much for chiming in.

You wrote:

So boot-arg slide=0 would be necessary to fix the random boot issues?
Have you tried any other BIOS? I tried 14xx and 1501.
At the moment:

• Z790 Creator, BIOS 1501
• Sonoma 14.1.2
• Opencore 0.9.7
• slide=0 boot-arg
• AirportItlwm works

By the way, I came across this post, where MrTrip wrote “turn off the password protection of EFI variables in the BIOS too so you can switch off cfg lock” but it doesn’t seem to be working, unless I am doing it wrong.

IIRC I was previously using 1501. I just tried reflashing it (coming from 1303) and it worked without slide=0; not sure what changed but it's booting now. However cxgb4 (Chelsio NIC driver) is still broken AFAICT; it loads but the interfaces don't seem to work properly. Shutdowns don't seem to be clean either; I get the "Your computer shut down because of a problem" warning after logging in (though I don't see any panic logs). Maybe UEFI runtime services are broken?

As for CFG Lock, I didn't have problems disabling it on BIOS 1303 after temporarily disabling that setting. I disabled it with:

setup_var.efi -n CpuSetup 0x44 0x0

The value persisted after the upgrade to 1501.

 

[m4rk0]

IIRC I was previously using 1501. I just tried reflashing it (coming from 1303) and it worked without slide=0

Apologies for insisting: are you saying that earlier, on 1501, without slide=0, it did not boot at all but it does now without it? If this is what you are saying, be aware my system used to boot 1501 without slide=0. But, at times, it just did not. Now I am on 1501 w/ slide=0, let's see if it is consistent. So far, so good.

Shutdowns don't seem to be clean either; I get the "Your computer shut down because of a problem" warning after logging in (though I don't see any panic logs).

I never got those on this (new) Sonoma install on a new Z790 Creator; by new I mean a few weeks old.
I had a few of those in the past, during my testings on the Z690 but I am not getting any on the Z790 (for now).
Also, I have this vague memory where using Onyx after getting those messages helped "cleaning" the system and getting rid of them. Try it?
If it's of any help, I can send you the BIOS profile and/on EFI folder.

PS just woke system from sleep, turned it off and turned it back on: no unpleasant message on my end.

As for CFG Lock [..]

setup_var.efi -n CpuSetup 0x44 0x0

The value persisted after the upgrade to 1501.

Wow it works! How you guys keep updated with all of these findings is beyond me.

@CaseySJ I was thinking to write a couple of lines which could be helpful for Z790 ProArt Creator users checking this thread... do you think it's okay to put a reference on the first post of this thread?
Also, IMO it would be good to include in the first page the firmware upgrade instructions for the onboard AQC113CS, as it fixed the wake from sleep issue and is useful for both Z690 & Z790 users - I flashed the Z790 as well, obviously

 

[CaseySJ]

...

@CaseySJ I was thinking to write a couple of lines which could be helpful for Z790 ProArt Creator users checking this thread... do you think it's okay to put a reference on the first post of this thread?

If you can create a post/guide that explains both the problem and solution, and which motherboard(s) it affects then it can certainly be a reference added to Post 1.

Also, IMO it would be good to include in the first page the firmware upgrade instructions for the onboard AQC113CS, as it fixed the wake from sleep issue and is useful for both Z690 & Z790 users - I flashed the Z790 as well, obviously

Good idea. I've added this under Troubleshooting / FAQs because (a) it may not affect everyone and (b) it's part of the troubleshooting process for waking from sleep.

 

[Azimuth1]

If anyone is running 64GB on their ProArt with XMP enabled, please let us know if macOS boots and whether any BIOS or OpenCore modifications were necessary.

Working with 16GBx4 modules G.Skill 6600Mhz ... no issues so far but still on the last Ventura MacOS

 

[racermaster]

Apologies for insisting: are you saying that earlier, on 1501, without slide=0, it did not boot at all but it does now without it? If this is what you are saying, be aware my system used to boot 1501 without slide=0. But, at times, it just did not. Now I am on 1501 w/ slide=0, let's see if it is consistent. So far, so good.

I am currently not using slide=0. Previously I needed it to boot. I suspect it's a KASLR issue like I said before. Manually overriding the KASLR slide (instead of letting EfiBoot/OpenCore pick a random one) should make this behavior more deterministic. I'm staying on 1303 as it doesn't have these issues.

 

[MacArthur]

TL;DR Summary:[/B]
I used @CaseySJ's comprehensive hardware shopping and build guide for build out and excellent installation guide and had a mostly event-free installation with O.C. 0.7.8 and macOS Montgomery. There were a couple small hiccups along the way, which is to be expected: I'll share the lessons learned.

LOVE it
You made my day with your build already working on this new "macOS Montgomery" !
Upcoming macOS15?
Real thanks for your input on this set-up !

 

[m4rk0]

I am currently not using slide=0. Previously I needed it to boot. I suspect it's a KASLR issue like I said before. Manually overriding the KASLR slide (instead of letting EfiBoot/OpenCore pick a random one) should make this behavior more deterministic. I'm staying on 1303 as it doesn't have these issues.

Thanks. I'll keep 1501 instead and check over the next week whether slide=0 fixes it. I'll report back.

 

[Azimuth1]

Hackindrom is not being updated for supporting OC 097? Sorry my fault, it's working with OC097 too!