Hacker News new | past | comments | ask | show | jobs | submit login
Intel BE200 WiFi 7 cards locked to 14th gen CPU and Z790 chipset (twitter.com/ghost_motley)
47 points by p_l on Nov 7, 2023 | hide | past | favorite | 26 comments



Classic Hacker News. Some guy posts likely incorrect random speculation, and the comments are full of even more speculation.

The truth is Intel makes two versions of most of their WiFi cards: the normal version, and a "low cost" version where some of the internal components are missing and those functions instead offloaded to the Intel chipset on the motherboard.

Sort of like the "WinModems" of WiFi cards.

If you want one that works, buy the correct version.


It sure looks to me like "Some guy" got the "random speculation" right in this case.

The link shows the BE200 as the "M.2: PCIe*, USB" type in the highlighted Intel Ark screenshot. In contrast to the "WinModem" version you're conflating this with, which should say "M.2: CNVio2" (or similar) like the AX210 and AX211 pair.

BE200: https://www.intel.com/content/www/us/en/products/sku/230078/...

AX210: https://www.intel.com/content/www/us/en/products/sku/204836/...

AX211: https://www.intel.com/content/www/us/en/products/sku/204837/...

The BE200 should be like the AX210. The "WinModem" variant should be BE201 like the AX211.

Edit: I guess I should say that I don't know if "may be locked via Firmware" is a reasonable conclusion, just that the correlation is present, and the assumption of interface compatibility (given an M.2 slot with PCIe and USB).


Why are we "assuming" interface compatibility? I have an Asus mobo that refuses to boot with a Western Digital SSD in the first M.2 slot, when and only when using a 13th-generation Core CPU, unless the Intel ME has been updated past a minimum revision. I think it is quite likely that PC platform firmware has huge tables of device quirks that are necessary to bring up the machine.


Classic Hacker News.

Posting includes exactly that information, explicitly ruling out that the card is "low cost" CNViO RF module according to specification provided by intel - yet a post will ignore that :)


To be fair, many people can no longer read Twitter threads. But yeah.


I think there is something going on, as the BE200 is E keyed rather than A+E keyed like the AX210. It physically won't fit in the slot in my AMD laptop where I previously installed an AX210.


What's also funny is they've been doing this for many years. I assumed it was common knowledge by now.


It's also explicitly discussed in the post and checked for.


To me its a bit surprising the platform wont even POST, seems a bit more like a problem with the platform than the PCIe card no?


No, I have a Camera Link PCIE card that causes any system I put it into not to boot. So it could be the card. But also, AMD firmware is...interesting... based on my experiences with Epyc motherboards.


Not really; link training and initialization for PCIe devices is really complex. A lot of motherboard bootware, even from reputable vendors, tends to handle obscure conditions by simply rebooting or locking up.

I wouldn't be surprised if there were certain steps where they simply forgot to include any sort of timeout, meaning the device can hang the boot process by simply not responding at the right moment.




So what pin do we need to tape over this time?


Conspiracy theory but it's pretty funny how CPUs would work for years while GPUs would get (rightfully so) outdated fairly fast.

Now suddenly CPUs do this weird force upgrade with obscure features, and even Microsoft forcing people to upgrade with stuff like TPM that the vast majority do not care about.

Wouldn't surprise me if Windows 12 next year announces that it will only support intel cpus from 12th gen and forward and ryzen 7000 and forward.

It might really be time to migrate to linux on my non work pc


Rando owns some dodgy PCs that won't POST with this card installed. Don't leap to conclusions.


works like a charm in M.2 wifi slot on i7-7700 w/ Asrock Z270M Extreme4


netsh wlan show drivers

Interface name: Wi-Fi

    Driver                    : Intel(R) Wi-Fi 7 BE200 320MHz
    Vendor                    : Intel Corporation
    Provider                  : Intel
    Date                      : 9/6/2023
    Version                   : 23.0.5.7
    INF file                  : oem75.inf
    Type                      : Native Wi-Fi Driver
    Radio types supported     : 802.11b 802.11g 802.11n 802.11a 802.11ac 802.11ax 802.11be
    FIPS 140 mode supported   : Yes
    802.11w Management Frame Protection supported : Yes
    Hosted network supported  : No
    Authentication and cipher supported in infrastructure mode:
                                Open             None
                                Open             WEP-40bit
                                Open             WEP-104bit
                                Open             WEP
                                WPA-Enterprise   TKIP
                                WPA-Enterprise   CCMP
                                WPA-Personal     TKIP
                                WPA-Personal     CCMP
                                WPA2-Enterprise  TKIP
                                WPA2-Enterprise  CCMP
                                WPA2-Personal    TKIP
                                WPA2-Personal    CCMP
                                Open             Vendor defined
                                WPA3-Personal    CCMP
                                Vendor defined   Vendor defined
                                WPA3-Enterprise 192 Bits GCMP-256
                                OWE              CCMP
                                WPA3-Enterprise  CCMP
                                WPA3-Enterprise  TKIP
    Number of supported bands : 3
                                2.4 GHz [ 2412 MHz - 2472 MHz]
                                5 GHz [ 5180 MHz - 5885 MHz]
                                6 GHz [ 5955 MHz - 7115 MHz]
    IHV service present       : Yes
    IHV adapter OUI           : [00 00 00], type: [00]
    IHV extensibility DLL path: C:\WINDOWS\System32\DriverStore\FileRepository\netwtw6e.inf_amd64_71f8441d5224dfab\IntelIHVRouter12.dll
    IHV UI extensibility ClSID: {00000000-0000-0000-0000-000000000000}
    IHV diagnostics CLSID     : {00000000-0000-0000-0000-000000000000}
    Wireless Display Supported: Yes (Graphics Driver: Yes, Wi-Fi Driver: Yes)


How does the wifi firmware know what cpu you're running?


How do you think that "cat /proc/cpuinfo" works ?

Edit: to be clear, the PCIe BIOS subsystem runs some firmware on the host CPU.


Damn, who could have seen this coming.


I have a variant of the machine he tested (he used the Ryzen/nVidia, I have the Ryzen/Radeon) this in and it’s one of the most dumpster fire things I’ve ever dealt with from a firmware perspective. It’s to the point that the community has developed replacement firmware for it just to make it run properly.

It would not surprise me in the slightest that it’s simply the firmware not playing nice with a chipset it never knew existed.


good good, i like seeing other companies copying the apple playbook.

maybe when the tech world gets enshittified enough we will finally start seeing some government regulation on these topics.


Non news item, most laptops' BIOS has a PCIe whitelist for wireless cards as a way to make sure FCC doesn't come after them for users retrofitting cell modems / arbitrary antennas. You can hack the BIOs to bypass the whitelist with a couple hours of work.


The whitelist seems to be on the network card side this time, not on the shitty laptop side.

It could just be bad firmware from Intel running into trouble if the hardware doesn't get the initialised right; I'd expect the card to simply not function in case of an intended whitelist, rather than refuse to let the system POST like shown in the Twitter thread.


That is NOT at all what the article talks about.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: