Framework Desktop Ryzen AI MAX 300 BIOS 3.04 Release BETA

Framework Desktop
,
1

Highlights

  1. Updated the AMD PI code to version 1.0.0.2.
  2. Added support for 3A charging profiles on USB-C ports, enabled only when the BIOS setting “Force power supply on in standby” is set to “Enabled.”
  3. Improved the PXE Boot behavior, ensuring the system automatically attempts the next boot device upon a network boot failure.
  4. Enhanced the Power On AC behavior, allowing the feature to work correctly without requiring the system to boot into the Operating System at least once for initialization.
  5. Fixed an issue where the system incorrectly showed an “invalid supervisor password” error when the user set the TPM to Hidden and set the supervisor password in the same BIOS session.
  6. Fixed an issue where the storage password could not be removed after being set in the BIOS.
  7. Resolved a bug preventing the specific keyboard from functioning while in the BIOS interface.

You can check your current BIOS version following the steps here to determine if you are on the latest release.

After the beta release, we will monitor community feedback, and publish this release to our stable release channel after approximately one week if no major issues are reported.

Subscribing to release notifications

If you want to subscribe to new release notifications you can now opt in through this link to receive an email when we release a new BIOS or driver update for your Framework Desktop.

Downloads

Windows

Download Link SHA256
Framework_Desktop_Ryzen_AI_MAX_300_BIOS_3.04.exe BC365DC789549C16700DB05B02CE9CAD83ECE2D8A023759C1E8C3596737E3E2D

Please do not interrupt power to your desktop while the update is in progress.

Instructions for Windows Installer:

  1. Run the .exe.
  2. Click yes to reboot.
  3. Wait for the firmware progress bar to complete, and then the system will reboot.

Linux/LVFS

Please do not interrupt power to your desktop while the update is in progress.

Updating via LVFS is available in the testing channel during the beta period.

You can enable updates from testing by running

fwupdmgr enable-remote lvfs-testing

Please note that you must update with a charger attached, then run::

fwupdmgr refresh --force

then

fwupdmgr get-updates

then

fwupdmgr update

Linux/Other/UEFI Shell update

Download Link SHA256
Framework_Desktop_Ryzen_AI_MAX_300_BIOS_3.04_EFI.zip B589ED7D853250FD9B5CCD9A41AC2F37D4050FE31025BE7402FA1BEC69C72254

Note that if you use the EFI shell update with Windows, you should suspend Bitlocker if enabled before updating using the EFI updater.

Please do not interrupt power to your desktop while the update is in progress.

Instructions for EFI shell update:

  1. Extract contents of zip folder to a FAT32 formatted USB drive. Cleanly unmount the drive before physically removing it, otherwise the BIOS update may not function correctly.
  2. Set the ESP flag for the FAT32 partition.
  3. Boot your system while pressing F12 and boot from the thumb drive.
  4. Let startup.nsh run automatically.
  5. Follow the instructions to install the update.

Security Fixes

N/A

Enhancements

  1. Updated the AMD PI code to version 1.0.0.2.
  2. Added support for 3A charging profiles on USB-C ports, enabled only when the BIOS setting “Force power supply on in standby” is set to “Enabled.”
  3. Improved the PXE Boot behavior, ensuring the system automatically attempts the next boot device upon a network boot failure.
  4. Enhanced the Power On AC behavior, allowing the feature to work correctly without requiring the system to boot into the Operating System at least once for initialization.

Fixes

  1. Fixed an issue where the system incorrectly showed an “invalid supervisor password” error when the user set the TPM to Hidden and set the supervisor password in the same BIOS session.
  2. Fixed an issue where the storage password could not be removed after being set in the BIOS.
  3. Resolved a bug preventing the specific keyboard from functioning while in the BIOS interface.

Component Versions

This BIOS update is a bundle of updates to multiple embedded components in the system.

Not all of them use the same version number.

Firmware Version Updated?
BIOS 3.04 Updated
EC ec_304_a4ef660 Updated
PD 0.0.0E Same
AMD PI StrixHaloPI-FP11 1.0.0.2 Updated

Reporting Issues

To report issues we have created a public issue tracker on github. GitHub · Where software is built We hope that this is a better way to track issues with community involvement moving forward as we have found it difficult to both gather relevant information about issues people are reporting on the forums, and track the issues through their lifecycle in a transparent way.
If you do experience an issue with the update that is related to your system firmware, please post as complete a description as you can, including relevant system information, and external peripherals. Please note that we do not currently have a SLA for responding to issues on github, but we will be reviewing them through the bios release process, and will review them for future updates as well.

If you have an issue regarding hardware, broken devices, returns, etc, this is not the place, please contact support.

Known Issues

N/A

1 Like
2

This seems to have resolved my USB4 external SSD problem, though I need to monitor for several hours to be sure. It usually failed on boot but occasionally hung on for a while.

And thanks for fixing Logi Bolt UEFI support! That was a neat surprise!! You guys rock!

Edit: can confirm, with the requisite UEFI setting enabled (which doesn’t affect me as I disable standby, causes too many problems) the drive is regularly operating in the 15W power domain (4-11W actual, 15W negotiated)

Edit 2: drive continues to remain powered on under extreme load and only throttles thermally now.

Edit 3: aaaaand…. SSD power issues recurring. That didn’t take long. So it can operate in the 15w envelope but doesn’t seem to be keeping it?

3

I went through the issues in GitHub and did not see any mention of the PSU fan. I think in the BIOS the idle draw is too high for some reason. Since I am just a layperson and don’t really know what I’m talking about I also don’t know how to report this as an actual issue. So, let me post this in the form of a question, are you all looking at adding PSU fan support in the BIOS?

4

Very sad to see this. Do you have any more details on what happened? Did it renegotiate to lower power or the hardware can’t provide enough power?

5

I had unplugged it to plug my power meter into it. In doing so, and from further testing, I have determined that if the drive is plugged in on boot, it grabs the 15W PD profile. If plugged in while running, it only grabs 7.5W. The PD firmware does not seem to handle hot-plug at 15W.

On boot:

PD15
PD151527×2028 206 KB

on hotplug (showing zero draw because the drive powered down due to power constraints):

PD7
PD71527×2028 239 KB

2 Likes
6

I see. Thanks for investigating it.

I understand this is a beta release but this is a trivial scenario that should have been tested in dev.

Unless there is some reason that I don’t understand, but it really shouldn’t be this complicated. The PD firmware should just advertise 15W all the time and call it a day (also no need for that UEFI config). Advertising 15W doesn’t mean the device will immediately draw 15W constantly. And most of the devices can’t negotiate 15W on demand anyway. Imagine what latency it would be for SSDs.

1 Like
7

I have confirmed that the 15 w profile only works if the device is plugged in when the desktop is started from a cold boot. Hot plugging a device results in 7 w consistently, and a reboot does not trigger the 15 w output either. It has to be a full shutdown and power on. So technically the 3 amp profile is properly advertising, but only for the first few seconds when the computer is first powered on

I’m not a firmware engineer but this honestly seems like it should be a pretty straightforward fix. It just seems like the scope of the 15 w advertisement isn’t configured correctly to extend to post-cold-boot scenarios

1 Like
8

Compared to the draw from the rest of the system, enabling a theoretical peak output of 15 w for two ports doesn’t seem like it’s that much of an additional ask. I also don’t understand why this is such a big deal. Unless I’m running a space heater from the port, it’s not going to consume 15 w continuously. And if I am using something that’s drawing 15 w continuously, I’m okay with the PSU fan spinning up. It’s honestly not as loud as people are making it out to be, my external SSD fan is much louder. But honestly 15 w is negligible compared to the 100+ w the APU can pull anyway. It’s not like we’re asking for full full PD 3.2 240w capabilities. We’re asking for 3 amps at 5 volts.

And yeah, why does standby idle have to be enabled for 15 w to be advertised? And why was there not a PD version upgrade in this bios release?

2 Likes
9

Totally agree. Exactly what I was thinking too.

It’s not the PD firmware’s job to enforce power saving.

The fix could have been simply hardcoding 15W and done.

10

Only issue i have seen is that my OS startup takes a much longer time now.

It is really strange, the machine boots to grub screen fine and when I select the boot entry it starts to boot but gets stuck for minute or two on the Booting XXXXXstate. after the wait it then proceeds to ask my LUKS password and rest of the boot is normal.

So not sure what is happening, all i know this started after I flashed this through fwupd. It might be something in my Fedora image too but this doesn’t happen on my FW13

11

When you think about it, the problem is actually the nvme ssd. My nvme can be told with a nvme command to limit its power draw to below 7.5W.
But i think this should be automatic. If usb negotiates 7.5W, the nvme should automatically limit itself.
There is probably a disconnect between the usb power negotiation and the nvme device so the infomation never gets through to the nvme device.
So, until the OS can be told what usb watts were negotiated, and thus send the correct nvme power limit command, i, by default, limit it to below 7.5W with an nvme command.
I get very reliable nvme ssd now.

12

Maybe, but none of the SSDs that one can buy work like that. Also FW laptops don’t have this issue.

Yep, but you just wasted your money on the SSD performance you are supposed to get.

1 Like
13

if you’re going to put brakes on the power consumption, just skip the hassle and plug it into one of the usb3 ports since you’re never going to hit USB4 transfer speeds with a PCIe 4.x NVMe at 7.5W. Technically a workaround but a total waste of capacity

1 Like
14

Something I don’t know if I can ask here, but would it be possible to change the power behaviour on the USBs?

In my case, it restarts a microphone on every regular reboot, unmuting it and turning back on its RGB (which I want off). It also causes an audible pop in my speakers because of that.

On my (ryzen 5950x) tower, this somehow was never an issue. Reboots would not cause that mic to lose and gain power back. Turning it off entirely would though, of course.

15

There are plenty of ssds that can do their full performance using less than 7.5W, even the currently fastest consumer pcie5 ssd (sn8100) can do that.

You only need a little over half the peak performance of a pcie4 ssd to saturate 40Gbit usb4 anyway.

Yes but most of that is in the pd controller firmware.

1 Like
16

@Quin_Chou Do you think that this beta firmware could help for sleep&wake TPM troubles listed here: TPM not working after longer uptime (Get-Tpm starts returning error 0x80284001) - #5 by sandie

I have seen that BIOS 3.04 brings updated AMD PI code. Could it help for the issue I am facing? Would you recommend upgrading to Beta firmware? When official release is expected?

I am away from FW Desktop PC, but after 10.12 plan to do further testing.