Hi,
Got it. How about the test with a limited DLA request?
Do you still reproduce the issue?
Thanks
Hi,
Got it. How about the test with a limited DLA request?
Do you still reproduce the issue?
Thanks
Hi.
We are currently testing with an increased number of devices.
One device exhibited symptoms after 4 days, while the remaining devices are operating without issues.
The device with the longest uptime has been running for 7 days.
Restricting concurrent DLA requests appears to reduce the frequency of these symptoms.
Is there a GPU fallback in the DLA model, and could that also be a factor?
Thanks.
Hi,
Unfortunately, we will need to reproduce this issue locally to know more details about the cause.
Could you try to add --verbose so it will output more details when the issue happens?
More, is it possible to upgrade to the latest software for your custom board?
Since there are lots of updates from Jetpack 5.1.2.
Thanks.