Can I just install the OS/JetPack on the NVME drive using the dev kit carrier board, then toss it into the Turning Pi 2 with the Orin Nano? Seems that it'd be much easier and won't require any special cable/ adapters using the instructions.

Yes. That's what I did with mine.

Much appreciated!

One thing to note: NVIDIA released a new version of JetPack late last month (August). It includes Linux for Tegra 35.4.1. I'm having a heck of a time trying to flash it successfully on either Xavier NX 16GB or Orin NX 16GB with an NVMe device on a legacy Jetson Nano carrier board. I'm using SDKMANAGER on the Xavier NX and flash.sh on the Orin NX. Both seem to flash Linux successfully, but neither has been able to boot properly. I'm using the NVIDIA Boot Manager for first boot. I'm trying to flash it to the eMMC on my Jetson Xavier NX 16GB module now. Jetson Orin NX/Nano do not have eMMC. I tried a USB 3.0 stick as well on the Orin NX with what appeared to be identical behavior. Output from the debug/UART port suggests that Linux is panicking immediately after bringing up the login prompt.

I just burned a week trying to do this very thing, install boot Jetson on NVME on my Orin Nano. I first tried from one of Cranky Cyborg's two Nvidia forum posts, one of which was to use the SD card to create a GPT partition and install there. Nope. I have a Windows machine and tried VMWare hosting Ubuntu 20.04 using SDK Manager and Jetson Hacks scripts. Nope. Then Cranky Cyborg's second forum post using his Virtual Box/Ubuntu/SDK to get better USB options. I was able to get Nvidia APX support there, but not Linux Tegra USB. I tried my Win10 Laptop in case it's usb-c port was better. Nope. In the forums, the Nvidia people said no virtual machines. In the end I wiped on old Dell Winbox I had and installed Ubuntu/SDK and flashed using a USB-A to USB-C cable. Worked like a charm. But after boot recovery mode, I took out the jumper and got nothing. Then I realized I had to take out the SD card to get it to boot after I flashed the NVME. Really annoying that they can't come up with some better way. Now I'm looking to get the Turing Pi 2. Came here to make sure that I could install it on the board without having to go through this pure hell again.

I swear NVIDIA changed something involving the default size of the root file system between L4T 5.1 update 1 and 5.2.1. The installation process for NVMe used to allocate all available space to the root file aystem. Now, it's the same size as a 16GB SD card. Using sdkmanager to install more than L4T itself and NVIDIA 's runtime products, overflows the root file ... silently. Having a TTL/UART to USB cable was invaluable.

Have you tried the method from our docs that's working on the TPi2?

I have cables strewn everywhere cuz of various posts about USB compatibility. I'm surprised by the partition of the nvme. I don't know what I expected. At this point just glad I finally got it done. Last night, last thing. I was almost scared to boot it. Glad I finally listened to Nvidia and found a way to direct connect, pc to Jetson. Jeez.

Also, are you using Orin Nano devkit board?

Actually, the TP help site got me going in the right direction with understanding the USB stuff. I thought it was super helpful to get me on the way.