I am running openSUSE Tumbleweed with KDE and today the firmware update showed up on Discover. Everything appeared to be working fine and pretty hands-off. Reboot, flashing of the new firmware until it said “complete” and went to a renewed reboot.
… except nothing was happening. I gave it some time but I guess got too impatient after about 5 min of nothing happening. I first just pressed the on button as it maybe just shut down but nothing happened, yet my keyboard remained unlit, so it was not booting either.
I know, its better to let it be for half an hour or longer instead of being sorry but I made a hard restart by capping power at the power plug. Following that after a short while it started to light up the keyboard but only for a few seconds than it got dark. This repeated two or three times until finally it booted for real and everything started up normally. Checking the firmware, the update appeared to be successful.
This was quite nerve racking as I had already once no boot issues with a presumably messed up boot loader where I could not even get into Bios. Was this just me being impatient and making the mistake of not giving it enough time? What should happen after flashing is completed? Is there are some extensive retraining time of the memory or something going on?
I am having some issues viewing the BIOS configuration on a Display Port connection. I had to resort to HDMI to be able to boot from USB, which is really annoying. Prior to the update, I don’t remember having any issues with this machine.
The update takes forever, and the RAM seems to go through training again after the update. yeah the whole process was nerve-wracking and the progress bar is tortuously slow, but patience is key.
My major gripe is that they didn’t follow through with the proper 3A PD profiles so I can’t use some of my USB4 peripherals without falling back to one of the USB3.x ports
This would be a problem because when connected to USB3.x ports, the ASM2464 based SSD enclosures don’t support SSD trim. Pretty much all the USB4 SSD enclosures you can buy are ASM2464 based.
That was my problem. It did not really reboot, but maybe that was really just an excessively long RAM training. Next time I’ll train my patience beforehand and just pull the plug after an hour of nothing visibly happening during what should be a reboot.
Mine took quite a while to reboot. There was a spinning Framework logo for a long time, then it rebooted again, which also took a long time. Didn’t time it as I just worked on something else while waiting. It was several minutes though.
Exactly, completely unacceptable. Not only do I get 25% of the speed, it interfaces as a disk interface rather than native NVMe like USB4 does. This 3A profile NEEDS to be fixed or the desktop is a failed experiment.
As I’m waiting for my Batch 14 or Batch 15 128Gig delivery, I have time to mull… So suddenly I find myself with a new project:
Of determining whether other Strix Halo systems I can buy have been tested and confirmed to offer full USB4 support/capabilities [3A, 40gbps, 256 byte packet size, et al], or if these issues are a Framework-only thing.