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

[gandem]

@CaseySJ,
Speaking of hope... any reason for hope on the Thunderbolt bus front?

 

[ffjigft]

Confirmation of what I said months ago, that with nvm36, intel unleashed a buggy mess for Maple Ridge. Glad Asus is aware of the problem. But it is unfair to us users that nvm36 wasn’t properly tested for QA ahead of its release or that Asus didn’t WARN users of the fact that BIOS 2204 will upgrade the nvm, and that the nvm has side effects like disabling support for TB1/2 devices. Or that in my case, with XMP enabled, the tb controller will crash when the system resumes from sleep (if a tb device is physically connected).

As this was a sanctioned bios update from Asus, whether unintentional or not, the net effect of a broken Thunderbolt controller is like malware… but not from a hacker but from the vendor.

The other question I have is: what about nvm38 for z790? In my testing, the same buggy mess that we experienced with z690 nvm36 is also present in nvm38. At least for z690 it appears we can downgrade back to nvm31. But what about z790? For those experiencing problems there’s no nvm to downgrade back to…

Some kind of conspiracy by myself:

1. Intel and Apple are contending for the leadership of Thunderbolt standard. Check that in M2 MacBook Air & Pro, Apple uses a new controller labeled "U09PY3" instead of Intel JHL 8040R, the last Intel chip in MBA and MBP13. Quiet clear that Apple want to withdraw from Intel's impact.

2. Intel deliberately damages device compatibility, forcing 3rd party device manufactors to follow Intel's standard, and forcing users to make choices (buy new device with new Intel certification & "Intel Tax").

Regarding professional audio interfaces & adda converters, most of them use Thunderbolt 2 and even firewire 400. Even for Thunderbolt 3 audio interfaces, the situation is still worrying. Among all the manufactors, only Universal Audio stated clear that their Thunderbolt 3 devices would support Thunderbolt 4 in Windows 10 & 11. Antelope audio warned compatibility problems officially. Focusrite devices have been confirmed not working with PC which has Thunderbolt controller embbeded in CPU (Very likely these devices would not work with the rest of Thunderbolt 4 PC with Intel's new firmware).

3. Asus and other motherboard manufactors are forced by Intel to push new Thunderbolt controlller firmware & ME firmware.

If I guessed right, Asus would not offer further support, provide tools for downgrading firmwares on new motherboards (like Z790), as this approach would go against Intel's expectation.

 

[dehjomz]

Some kind of conspiracy by myself:

1. Intel and Apple are contending for the leadership of Thunderbolt standard. Check that in M2 MacBook Air & Pro, Apple uses a new controller labeled "U09PY3" instead of Intel JHL 8040R, the last Intel chip in MBA and MBP13. Quiet clear that Apple want to withdraw from Intel's impact.

2. Intel deliberately damages device compatibility, forcing 3rd party device manufactors to follow Intel's standard, and forcing users to make choices (buy new device with new Intel certification & "Intel Tax").



3. Asus and other motherboard manufactors are forced by Intel to push new Thunderbolt controlller firmware & ME firmware.

If I guessed right, Asus would not offer further support, provide tools for downgrading firmwares on new motherboards (like Z790), as this approach would go against Intel's expectation.

If intel can’t implement Tthunderbolt properly, then perhaps it is time to sunset it and turn development over to a different entity. Titan Ridge worked well enough but Maple Ridge has been very problematic.

Perhaps alternate USB4 implementations from non-Intel vendors might result in improved compatibility and stability.

 

[etorix]

The pack (in Google drive and the copy in attach here) includes 3 files ("ProArt-Z690-CREATOR-WIFI-ASUS-9901" in .cab, .cap, .rom). Strangely the version is 9901 (also shown in my BIOS dashboard after flashing). I'm not sure whether I can flash to other BIOS afterwards (during contact, I mentioned that my current BIOS is 1720 with ME 1735. I don't know if the BIOS is a modification of 1720 or not). As far as I tested, the .cap file works no problem with EZ Flash, my Thunderbolt firmware went back to NVM 31, and my Focusrite interface went back alive! For me, I probably would not attempt to update BIOS and ME afterwards (in order to mantain compatibility). Regaring Asus HK, I could not appreciate more for their diligence and helpful service.

Welcome to the forum, and thanks for a very valuable contribution! You set a high bar for first post. This is a very important warning for all users of old Thunderbolt 1/2 devices with any motherboard featuring Maple Ridge—not just Asus.

The other question I have is: what about nvm38 for z790? In my testing, the same buggy mess that we experienced with z690 nvm36 is also present in nvm38. At least for z690 it appears we can downgrade back to nvm31. But what about z790? For those experiencing problems there’s no nvm to downgrade back to…

Based on this report, you could contact Asus, explain your issue… and hope you'll get a custom NVM-downgrade BIOS for the Z790 ProArt.
Or maybe a sorcerer (@Elias64Fr ? @CaseySJ ?) can extract just the code to downgrade to NVM31 from the 9901 BIOS?

 

[CaseySJ]

Added to Post 1:

 

[Azimuth1]

If intel can’t implement Tthunderbolt properly, then perhaps it is time to sunset it and turn development over to a different entity. Titan Ridge worked well enough but Maple Ridge has been very problematic.

Perhaps alternate USB4 implementations from non-Intel vendors might result in improved compatibility and stability.

Uhm... it's not so simple... Asus Titan Ridge EX3 is one of the "hardest" card to flash with modded firmware, and the only viable solution it's been CG Titan Ridge so far. Which is - to my little understanding - the last one which ensures full compatibility with Mac's implementation, if flashed.

Asus Ex4, which is Maple Ridge, doesn't have any modded firmware just because Apple is not using Intel's chips anymore, and I really don't know any other vendors in the Intel world - where we live - that have some USB4 card with full Thunderbolt capabilities.

 

[Azimuth1]

Added to Post 1:

View attachment 564641

Will you try it on your machine? I was thinking that - just maybe - if it's a special bios perhaps has to possibility to be installed even in 2xxx systems, taking them backwards where the compatibility issues are far less in general and not only regarding the Thunderbold devices.

 

[etorix]

Some kind of conspiracy by myself:

1. If I'm not mistaken, Apple's Thunderbolt controller is in the SoC. The achievement of developing their own controller may speak of the desire to integrate everything into the SoC as much as the will to remove dependency from Intel.

2. Planned obsolescence is indeed a plausible explanation. A more benign option would be that Intel simply wants to reduce its development and validation costs by eliminating testing with older devices.

3. I suspect that manufacturers just push through whatever updated code comes from Intel.
If so, there's still hope that complaints may be answered by options to downgrade.

Perhaps alternate USB4 implementations from non-Intel vendors might result in improved compatibility and stability.

How many implementations are there, beside Intel's and Apple's?
If Intel were to drop development, I'd rather expect Thunderbolt to quickly fall into oblivion.

 

[dehjomz]

1. If I'm not mistaken, Apple's Thunderbolt controller is in the SoC. The achievement of developing their own controller may speak of the desire to integrate everything into the SoC as much as the will to remove dependency from Intel.

2. Planned obsolescence is indeed a plausible explanation. A more benign option would be that Intel simply wants to reduce its development and validation costs by eliminating testing with older devices.

3. I suspect that manufacturers just push through whatever updated code comes from Intel.
If so, there's still hope that complaints may be answered by options to downgrade.


How many implementations are there, beside Intel's and Apple's?
If Intel were to drop development, I'd rather expect Thunderbolt to quickly fall into oblivion.

I want Intel to win. I've been rooting for it to come back.

But TB4 on z590 had all sorts of ho tplug and compatibility problems. Which were never resolved.

And while z690 nvm31 was mostly fine initially (except for hot plugging not working for jhl6240 devices), nvm36 is causing major widespread issues according to my ongoing technical support with Asus. No resolution yet. These are production nvms with major problems. Ruining user experiences. Causing plenty of frustration.

My attitude is if you’re going to be a steward of a technology like Thunderbolt, then, in my opinion, you have to do serious QA. With nvm36, since the tb controller dies during sleep/wake, I’ve lost an nvme drive. SMART errors. First ever nvme drive I’ve ever lost. Luckily it was a relatively empty test drive. But it could’ve been a drive with terabytes of data including source code or music. I know we ought to have backups yes and I do, but this is a severe issue in my opinion. Why did intel feel the need to release nvm36?

Asmedia has been rumored to be releasing the ASM4242 USB4 controller.

 

[ffk]

Why did intel feel the need to release nvm36

I’m not exactly sure what I started on when I bought this board, but I am booted in Windows now and just checked. I am now showing NVM 38.0 on Z790 ProArt.