Labels for Port Capabilities

Not sure I use windows mainly. But the label-maker works on its own too.

https://apz.fi/blabel/

3 Likes

As a Batch 8er still dont have device but had 30 seconds to design my labels i have a D600 so can do 0.96" label here are my ideas
ok I was board and did pictures (I used generic pictures then Cleaned them up for a 180 DPI printer.
V4

V3

Summary

Port 1 & 4
Screenshot 2024-03-20 174829
Port 2
Screenshot 2024-03-20 174738
Port 5
Screenshot 2024-03-20 174801
Port 3 & 6
Screenshot 2024-03-20 174522

I am working on assumptions here
1 all can work with A ports
! means High power A ports
240 missing means no 240 lol
Thaks to @truehd for common sense i was over thinking it…

2 Likes

I think it’s clearer if “Display” only appears by ports that support it rather than putting “No Display,” but whatever works for you!

1 Like

yeah you got me there LOL didnt think about that lol good thing im just playing around with it.

Just fyi, “USB 3.2” is not saying anything relevant. Customers and customer-facing specifications should really avoid the versions of specs, only the family is important and with how USB-IF intended their redesign, even that is mostly optional.

So USB3 in contrast to USB4 may be important.
If you want to denote the speed then USB3 10G(bps, GBit/s) would indicate the speed supported.
Since no FW device so far supports USB3 20G one could even use the official USB names and logos, because they are unique if one never considers USB3 20G: USB3 10G (= USB 10G, by definition included in USB 40G)

So if one wanted to list port capabilities it would be more like
USB4 Ports: “USB 40G, DP HBR3, PD< 240W”
DP/USB3 Port: “USB 10G, DP HBR3, PD< 240W”
Pure USB3 Ports “USB 10G”.

The warning about legacy USB equipment incurring higher power consumption can be listed in addition. Technically that does not just affect USB-A but anything that is not natively USB-C. So standard-compliant USB-C to microUSB cables would also be considered USB-C legacy adapters that are affected the same way.

One specification that also fell by the way side is power output.
TB4 mandates 15W power output as supported by the Intel FWs. Although technically that is not USB-PD but “Type-C Current”, so maybe one should not write this as “PD”.
USB4 actually does not mandate 15W, only 7.5W. If FW supports 15W (would be my hope and would match the Intel boards and parity with TB4), it really should be listed explicitly.
My suggestion would be matching the above scheme with “>15W”.

Especially for non-USB4 ports supporting 15W would also be relatively rare, so should be explicitly mentioned if supported.

While AMD specs the CPUs to support DP at higher UHBR10 speeds, the Kandou ReTimers FW lists in their schematics are only specced for max. HBR3.

All of this does only list what is directly supported on the port. What is supported through a USB4 connection could be/is different and should also be listed by full specifications (although might not be needed on the device). Neither AMD nor FW are doing that so it would require someone to get clarity on the exact tunnel support.
Looks to be like (USB3 10G, 2x DP HBR3, max. PCIe) though.

The 2x DP can be important depending on the dock and has only been confirmed experimentally by a few. PCIe speed and support is convention for most CPU-integrated USB4 controllers and mandated by Windows. But the 11th FW for example only supports minimum mandated by TB4: 32G.

Also, maybe something else FW16 owners would want to remember: all 3 non-DP-capable USB ports come off a single USB3 10G hub, so should strictly share bandwidth.

And yes, all of those specs & details should really have been listed by FW.

3 Likes

Now this is great feedback! I think the graphics provided by FW includes only the most relevant at-a-glance info that the average user would need to know before trying to connect cards. However, I do think there is value in being more verbose for users who who need that info.

1 Like


So here is the final product.

6 Likes

I like it. Now if only the background color matched the laptop or was clear. That’s not a big deal though since it’ll be covered most of the time. That top left one being crooked bothers me though. :rofl:

yeah well it was the first one i applied it kinda bothers me but in all honesty will only bother me when/if i need to replace the module lol…i used the labels i had. and as you said its covered up most of the time.

The DP/USB3 Port however uses a ANX7493 redriver, which is listed as having support “for DisplayPort up to UHBR10”. (Source)

I haven’t seen any confirmation regarding if that port supports UHBR10 or not.

Yep. I don’t think any notebook on the market, Intel or AMD has ever advertised higher than HBR3 speeds for the iGPU, even though the CPUs are supposed to have supported it for a few generations.
If most current AMD notebooks had UHBR10, I’d use that chip as confirmation that it is supported here as well…

Btw, I have found another specs page of the Kandou ReTimers that says UHBR20 (10 and 20 at least would make sense as that is basically just USB4 signaling with all 4 lanes in 1 direction). So there might be firmware updates or revisions of that chip that also would support UHBR10.

But also, we have seen that AMD’s public specs are not that reliable (and they are not detailed, so they are not what is actually used for development & manufacturers). After all, with the ECC DRAM thing, FW was questioned a lot if that would be sth. that is possible. And shortly after FW told us it was not supported by AMD, AMD changed the public specs to also say the same, after they had been up for a long time. So I kind of do not trust just those public specs, if no manufacturer has ever implemented it.
That might just be marketing for AMD, but not actually “supported” for any manufacturer for whatever reason.

1 Like

I also noticed that on Framework’s schematics they do list “DP2.0”. To me that implies at least UHBR10, although I don’t know if DisplayPort 2.0 actually requires that or if it can be implemented with just HBR3.

Also, TechPowerUp seems to claim that the processors only support DP 1.4a over the USB4 ports.

Overall there seems to be a lot of conflicting data. The only way to know for sure is probably for someone to plug it into a monitor that supports UHBR10 or greater and see what it runs at.

It does not. As with most other protocols. DP 2.0 does not guarantee almost anything. Hence it should not be used to advertise a port (unless explicitly accompanied by “full support of all DP 2.0 features”).

Vesa’s stance has been a bit more nuanced, where they say manufacturers may only advertise / spec the number, if at least any feature that was added with that version is supported. But you probably could not count the schematic of the board as “advertising”.

Whereas the other standards (HDMI, USB for example) take more the stance, the newest version replaces the older versions completely and immediately, the older version does not exist and so none of the features are required. Ports simply have to follow the newest specification no matter what they support. And it is forbidden to use the spec number as if it meant anything (though most manufacturers violate that heavily).

I.e. the “USB 3.2” right next to the DP 2.0, by its very definition is pretty worthless. And indicates only that it is USB3, which can range from 5 Gbps to 20 Gbps. It is basically only an acknowledgement that the board was designed after USB 3.2 came out and looking at the current PDF file while designing it.

1 Like

I’ve made some labels in Onshape (when you have a hammer…). I’m picking up some laser label paper, and will test them out once I have my framework.

edit: A few tweaks,

PDF of Labels
OnShape Model
Label Paper

2 Likes

You can attach thunderbolt devices to the usb4 ports and they will work.
I have 2 different thunderbolt devices and they both work in the AMD FW16, but only in port 1 and 4.

It depends on the device and which specific thunderbolt features it uses. USB4 is an open standard of Thunderbolt 3, where most features are optional (and without Intel’s certification requirements.)

AMD, so far has pretty much kept pace with Intel’s abilities / TB4 requirements. And because of Windows’ requirements any USB4 host capable of running Windows will be TB3 compatible and have PCIe tunneling for the foreseeable future. Only amount and capability of DP connections and PCIe bandwidth could be less. Even Qualcomm has 2 HBR3 DP tunnels. I just don’t know of their PCIe bandwidth.
Asmedia for their own Host controller has matched / surpassed TB4 requirements.

There is just no name for that label of functionality that is not tied to Intel and its certification. I often say “TB4 level”, but that is not an option for manufacturers.
Will be interesting to see future controllers. Because Intel is in the progress to upgrade to 3 DP tunnels (with TB5, but also with new TB4 controllers). So anybody that wants to make use of new TB hubs will also need those 3 tunnels.
But with Intel leaving out the middle DP UHBR speed, there is also more room to differentiate…

I got my framework and did a small arts and crafts project.


I will probably reprint port 1s sticker, since it got pulled off and reapplied a few times.

3 Likes

3 and 6 ports only supply up to 4.5W and may not compatible to all USB 3.2 devices, how about adding the power rating there as well