Hello Thank you for the reply. Is this something that will be fixed in the future ? Is it a firmware fix or a hw fix ? The UHS-I mode card I/O throughput is pretty low. Is there a workaround for this ? Can I have all the filesystems except /boot on a USB-3 connected device ? Will this improve performance ? What is nVidia’s recommendation for this ?
ERROR, WARNING, and FAILURE messages in logfiles should be cleaned up before releasing to the public.
What if Nvidia produced a patched and usable system image every day, like famous Linux distros?
Because all of this is hard. So it’s not perfect. Not many people have Jetson Nanos yet, and most of these issues are spcific to Nano.
As a result not meny people have tested. As a result not as many people have reported. As a result the developers don’t find the bugs that are there.
As a result the bugs don’t get fixed. Solution? Bug reporting. If you find an issue, report it… or you could just wait and the platform will stablize on it’s own.
Yes, there are issues. Never has a release of a hardware of software product gone without them. OTOH most stuff works and you can talk to (some of) the people who actualy write the software, ask them for advice, etc.
RE: UHS speeds, I saw this documented somewhere in your documentation but I can’t find a link. I think it was somwhere near the pci lane configuration documentation for Nano but Niether google or your internal search is yielding any results. Any idea where I saw it?
The answer is right there: It is possible to use an UHS-II card, but not at UHS-II speeds.
If this is not acceptable to you, don’t use a Jetson Nano.
Yes it’s backwards combatible, that’s not the question. See how Manjunath asked a good question, but noone bothers to answer. Always easier to say don’t use it.
What’s the point of dev talk if those who have schematics, can’t lookup if reader has UHS-II pins or nah.
I had to search for a while and found, that it doesn’t.