Intel has released Wifi 7 cards: Intel® Wi-Fi 7 Series Products and Solutions with Intel® Wi-Fi 7...
Are these compatible with Framework 16? If yes, wouldn’t it make sense to add the BE200 to the configuration options, or better make the default, for being future-proof and seemingly sold at a good price? (The BE202 is just slower but almost same price)
If is Framework planning to release firmware updates on that?
Given it doesn’t work on the amd 13 at all and they share the same platform there’s a good chance it won’t work on the 16 either.
Would be neat if it did though and it is possible the incompatibility will be fixed in a future agesa update or something.
Thanks for the Info!
For a feature request, should I send a request to support, or will this post seen and forwarded by framework staff here anyway?
Lots of intel WiFi cards use the CPU for certain tasks, so guess this will never work on AMD.
There is an wifi7 from mediatek, that wil work.
I don’t think framework themselves can do much about that (other than not taking an eternity to release a new bios once fixed agesa is out)
agesa is entirely in the gands of amd.
The be200 is a pcie wifi card not the xx1 proprietary interface kind and it does apparently work in some very old intel platforms but at least in the case of the framework 13 it causes the system to not post and if connected after post (vie thunderbolt) it hard crashes the system (no kernel panic or anything fancy like that just off).
Thanks for the info. Quick googling didn’t help to figure out the model name.
Do you by change know the Mediatek Wifi 7 card’s model name?
Unfortunately no, still a new card. Not in alot of shops, but did search for it once but didn’t bookmark the page.
The MediaTek/AMD partnership Wi-Fi 7 card has the model number RZ738, however it is pretty hard to find.
Hi,
It has been mentioned that the BE200 does not post on the FW13 and AMD motherboard., so a guess might be that it will similarly fail post on the FW16.
The FW13 with 13th Gen Intel motherboard works with the BE200.
The reason the BE200 does not work on the AMD motherboard might be due to a “BIOS Whitelist”. I don’t know if “BIOS whitelist” is factually correct, but it is mentioned here:
https://forums.lenovo.com/t5/Gaming-Laptops/Legion-7-AMD-upgrading-to-WiFi7-Intel-BE200/m-p/5270899
be200 is a M.2 Type E card, and the slot on the Framework AMD motherboard is a M.2 type E slot, so the hardware should be compatible.
Given that the symptoms on the Framework 13 are very similar to the symptoms in the Lenovo (URL above).
Maybe the Framework AMD motherboard BIOS has such a “BIOS whitelist”. In which case, if Framework are so inclined (have enough time on their hands, given all the other higher priority items they are dealing with at the moment) a BIOS update is likely to fix the problem. Probably a first step to that, is someone to publish what is on the current “BIOS whitelist”. Maybe a second step is for Framework to remove the “BIOS Whitelist” so users have the freedom to plug in any future, better wifi card they wish to use.
So, if it is a “BIOS whitelist” and the BIOS on the FW16 is different from the FW13. The be200 may or may not work depending on what is on the BIOS whitelist, but it is most likely to be the same “BIOS whitelist” on the FW13 and FW16.
Another URL covering this “BIOS whitelist” topic is:
I doubt framework were even aware of such a “BIOS whitelist” feature so have probably not even taken a conscious decision on whether to have a “BIOS whitelist” or not.
My guess is that framework have taken a common, off the shelf (COTS) BIOS, and tweaked it just enough to make it work with the FW13 and FW16 and left it at that. They have probably never needed to delve deeper to every possible tweak option and made a choice on each and every one. It would just take too long to do.
I highly doubt this is a whitelist. Not only is whitelisting something I don’t think Framework would do, but whitelists are something that are applied boot and ignored afterwards. For instance people have been able to defeat Lenovo whitelists by somehow keeping the device connected to the wifi/wan slot powered off until after POST, and then powering it on, at which point the operating system would be able to see/use the device.
In this case, I’ve read that people have also experimented with connecting the BE200 wifi card in a PCIe to Wi-Fi card adapter in an eGPU enclosure, and the machine would immediately crash when plugging in the eGPU enclosure. That’s not something the bios is doing.
My hunch is that Intel is doing something non-standard over the PCIe bus, which AMD CPUs don’t understand, and therefore are crashing.
There is of course other possible causes. For example, the schematic for the FW16 shows a link between the Wifi socket and the EC chip of the wifi connector pins 38,40. Maybe a feature or bug in the EC firmware needs fixing to resolve this. As pins 38,40 are normally “vendor specific” pins, it is reasonable to believe that a different wifi card might use those pins for different things, and thus maybe upset the EC firmware and cause power problems.
One could test this theory. One can get M.2 M-type to E-Type adaptors. One could try the BE200 in the adaptor and put it in the NVME slot on the FW16 and see if it gets through BIOS post. One could also experiment plugging the BE200 into desktop AMD motherboard M.2 slots with the adaptor, so as to avoid any possible damage to the expensive FW16 main board.
My understand is these intel Wi-Fi 7 cards all appear to require CNVI which requires intel chipsets that have controller parts in the chipset itself so AMD would have to have the same controller equipment i would assume on their chipset.
Another problem might be the SDIO RESET pin. The FW16 schematic does not have this connected to anything.
I have seen problems with some old PCI-Express IO cards that are supposed to be normal PCI interfaces, but they would immediately throw PCI bus errors if you tried to contact them before first sending them a special control pattern. So maybe the SDIO RESET pin is used to control the power on process for the card. In which case, the BE200 is never going to work with the FW16 motherboard as the SDIO RESET pin is not even connected to anything.
Hi,
As a general comment, the firmware for the BE200 appears to be quite buggy, so I doubt it is going to improve a user’s wifi experience due to it probably having problems connecting and staying connected etc. So, I would probably avoid the BE200 altogether at this stage, even if you have an Intel motherboard and the BE200 appears to be recognized.
Msi has a WiFi 7 PCIe network card with a Qualcomm WiFi 7 module (ncm865). I ordered it to salvage the WiFi module and use it in my laptop.
Let us know how it works I’m scared to try Qualcomm after the disaster of a card I had in my XPS 15 (9575) 2-in-1 the qc1435 (After seven months of crappy WiFi was able to get an intel 8265 io board from the precision 5530 2in1 from support. (Was the same computer just had the intel with v pro for the business line)
Also at work we use dell latitude 7389(retiring thank god)/7320/7420s that all have Qualcomm cell card And they by far have more issues (Disappearing, dropping Sims randomly Having to shut the computer down, turning it back on in order for the card to just show back up) than the Intel cell cards in our 5300s and 5330 and the precision laptops we have all have intel cell(Various models)
You may wish to contact @GreyXor who has already had the card running in their laptop for a little while now.
I’m now using the QCNCM865 card. And it’s working like a charm with this ath12k(driver) version. kernel/git/kvalo/ath.git - wireless ath drivers tree, ath9k, ath10k, wil6210 etc.
this one is important for now because it contain some important fixes. All fixed should be mainlined in 6.9.
please AMA about the QCNCM865. I confirm that, here, it’s working very good.
When connecting the card via thunderbolt those pins aren’t connected to the ec and the laptop still crashes so I think that’s not it. Given that it seems to behave like this on a whole bunch of other mobile and desktop amd platforms it is likely the cpu/chipset really not liking something the be200 does, if the be200 is doing something sketchy (and just gets away with it on other platforms) or the amd platform is overly sensitive is a whole other question.