[ho1Aetoo]

so de facto all AC WLAN devices have modulators and demodulators for 256-QAM
Because IEEE 802.11ac requires 256-QAM - without this there is no certification.

Just as ALL AX WLAN devices have modulators and demodulators for 1024-QAM because the standard requires it.

However, this does not mean that the respective drivers support off spec usage in older standards.

Under the following links you can see why e.g. the ath10k (atheros AC) driver does not officially support 256-QAM in IEEE 802.11n.

https://www.mail-archive.com/ath10k@lists.infradead.org/msg04664.html

Because it is just off spec nonsense that is not widely implemented.
But this can be fixed with a small patch
Guess why the Atheros AC routers with dd-wrt support "turboQAM".
Someone simply added the necessary patch.

And that any ASUS cards use "add-on" modulators I may doubt.

They will simply use a Broadcom Wifi5 chip that comes with the appropriate modulators/demodulators.

See:
https://www.broadcom.com/products/wireless/wireless-lan-infrastructure/bcm4366

http://en.techinfodepot.shoutwiki.com/wiki/ASUS_PCE-AC88

Oh, and regarding the higher data rates, they only work at short distance and very good signal.

Otherwise, the device falls back to a lower modulation, see

https://mcsindex.com/

[the-joker]

Its still closed source tech, you can patch atheros drivers until cows come home and be free from any legal repercussions if the code is less than 20% similar to closed source solution (this was the established legal precedent in US law when IBM bioses where reversed engineered/cloned by Compaq in their early days)

Doubt all you like closed source is closed source and for full support requires fees being paid to benefit from that extra boost even if todays radios support the off spec modulation without extra HW (we wont know until datasheets are examined)

Not a simple open and shut door, especially where Broadcom has not owned the patents/licenses or tech since all that shite was sold to some other company.

Asus idk for how long they were licensed to be exclusive to this tech initially, clearly thats time has come and gone otherwise IEEE wouldn't have access unless they paid for the licenses, all I knw is initially it was separate HW than regular radios.

[ho1Aetoo]

Well the kernel developers see no problem in adding e.g. 256-QAM support for ath10k

Besides, the chip manufacturer has to pay license fees and the customers may not care.

All the manufacturers also advertise their devices with "256-QAM" support.

For example, the Netgear r7800 with AC2600 class - the 2600Mbit is composed of 1733Mbit AC and 800Mbit N.
And the 800Mbit N can only be achieved with 4x4 MIMO + 256-QAM.
The factory firmware also supports 256-QAM in N mode.

The only one that is really protected is TurboQAM™

[the-joker]

Trademark names are not the same as patent license fees for a specific tech. Don't confuse potatoes with onions.

And support may be added to opensource as long as code stays below a specific magic percentage number in similarity line for line. But opensource has a long history here and there of borderline implementations of patented technology. GCC compiler being one of such and H.264 and many others. But that's a different discussion.

In the end, this sausage factory went nowhere, patented technologies are closed source and full support for opensource drivers will not be easily implemented, name it whatever you like.

The point being is still not many wide spread client HW supports it fully and the ones that do are old shit Asus did and other companies did for corporate/military applications and scraps to the user base as usual.