Questions regarding MAXN availability for NVME only systems (no sd card)

[robotunderpants]

According to a previous post, the MAXN mode is not available at this time for non sd card Jetson Nanos.

Please check if your Orin Nano module has SD card slot(which is sku 5). Currently the mode is enabled for sku 5.

My model information is below. I have no SD card slot

cat /etc/nv_boot_control.conf
TNSPEC 3767-300-0004-N.2-1-1-jetson-orin-nano-devkit-
COMPATIBLE_SPEC 3767--0004--1--jetson-orin-nano-devkit-
TEGRA_BOOT_STORAGE nvme0n1
TEGRA_CHIPID 0x23

According to the linked post above Orin Nano modules without the SD slot are sku 3

Please check if the Orin Nano module has SD card slot. With the SD card slot, it is sku 5. Without it, it is sku 3.

However my sku appears to be neither 3 or 5, but rather 4?

So what exactly is my model and when will MAXN mode be available?

[WayneWWW]

Sku4 is Orin Nano 4GB. Previous discussion about Orin Nano sku3 or 5 are for 8GB.

You have to wait for next release.

[robotunderpants]

Am I correct that MAXN mode is now available in Jetpack 6.2 (Jetson Linux 36.4.3)?

Per https://docs.nvidia.com/jetson/archives/r36.4.3/ReleaseNotes/Jetson_Linux_Release_Notes_r36.4.3.pdf

Support for new reference power modes on Jetson Orin Nano and Jetson Orin NX
production modules, delivering up to 2x generative AI performance available with a
new ashing conguration: jetson-orin-nano-devkit-super.conf).
○ NVIDIA Jetson Orin Nano 4GB: Supports 10W, 25W, and MAXN SUPER.

[WayneWWW]

Correct. Please give a try.

[robotunderpants]

Am I looking in the wrong place

~$ sudo nvpmodel -p --verbose | grep POWER_MODEL
NVPM VERB: POWER_MODEL: ID=0 NAME=10W
NVPM VERB: POWER_MODEL: ID=1 NAME=7W_AI
NVPM VERB: POWER_MODEL: ID=2 NAME=7W_CPU

And should I see two jetpack revisions installed?

sudo apt-cache show nvidia-jetpack | grep Source
Source: nvidia-jetpack (6.2)
Source: nvidia-jetpack (6.1)

Looks like I have the correct Jetson OS verion and Kernel

$ cat  /etc/nv_tegra_release
# R36 (release), REVISION: 4.3, GCID: 38968081, BOARD: generic, EABI: aarch64, DATE: Wed Jan  8 01:49:37 UTC 2025
# KERNEL_VARIANT: oot
TARGET_USERSPACE_LIB_DIR=nvidia
TARGET_USERSPACE_LIB_DIR_PATH=usr/lib/aarch64-linux-gnu/nvidia

$ sudo uname -a
Linux jetson.lab 5.15.148-tegra #1 SMP PREEMPT Tue Jan 7 17:14:38 PST 2025 aarch64 aarch64 aarch64 GNU/Linux

[WayneWWW]

How did you flash the board? I guess the flash command is not correct.

Sdkmanager by default won’t install super mode config for you. You have to flash manually.

[robotunderpants]

I flashed with SDK manager on Ubuntu 22.04. Its very much news to me that I cannot use SDK manager flash process to get the correct image? I’ve not seen this mentioned before.

How do I flash it manually. I have only seen an SD card image, is there supposed to be an NVMe image, or should I just use the SD card image.

Can you point me to the correct image for the 4GB NVME and directions on how to flash without SDK Manager. Having a 128GB NVME I want to make sure that I have instructions on how to expand filesystems

[WayneWWW]

Hi,

Sdkmanager already downloaded the necessary files in your host PC with default path ~/nvidia/.
There would be a Linux_for_Tegra directory there and you shall run command from that path.

You need to put the board into recovery mode and follow this document to flash.

https://docs.nvidia.com/jetson/archives/r36.4.3/DeveloperGuide/IN/QuickStart.html

→

$ sudo ./tools/kernel_flash/l4t_initrd_flash.sh --external-device nvme0n1p1
-c tools/kernel_flash/flash_l4t_t234_nvme.xml -p “-c bootloader/generic/cfg/flash_t234_qspi.xml”
–showlogs --network usb0 jetson-orin-nano-devkit-super internal

[robotunderpants]

Thanks, I will give this a try shortly and let you know if I run into issues.

[robotunderpants]

Thanks that worked.