Hello everyone, NVIDIA’s happy to announce availability of standalone Jetson TX2 modules in several regions starting today. To purchase go to:
I bought the Jetson Tx2 Dev kit before the individual mdoules were avialable. I want to use them on a small drone with the orbitty carrier. Is it possible for me to extract the modules from the dev kit and use them . I am aware that this will invalidate the waranty but my question is will they work if I take the modules out from the dev kit.
Yes, the module can be unmounted from the devkit motherboard.
anynoe creat car type robot?
dusty_nv has one! It’s called Turbo2:
I’m also working on one, but it’s not up and running yet.
If you’re considering full-scale cars, then the people who work on those applications typically use the NVIDIA Drive series (PX2) rather than the Jetson. There are so many things that need to be done in parallel for safe driving, that you need more parallel oomph, and the car integration space has more stringent requirements as well.
I have a TX1 Dev Kit. Can I replace the Jetson TX1 module on the TX1 Dev Kit base with a Jetson TX2 Module?
Hi JonMorris, they share same carrier board as you can find the related design doc here: http://developer.nvidia.com/embedded/dlc/jetson-tx1-tx2-developer-kit-carrier-board-design-files
I recently got a Stand alone TX2 Module, is there any tutorial to get started with it?
Is there a preflashed OS on it as well as on the Dev Kit edition?
If yes, I did nothing special than plug it on my Orbitty Carrier and got no display (HDMI) and keyboard’s lokck/unlock key LED does not light up.
Thanks by advance.
What Ethernet chip is used on this SOM?
Here’s the ethtool output:
$ sudo ethtool eth0 Settings for eth0: Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Supported pause frame use: Symmetric Receive-only Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Advertised pause frame use: Symmetric Receive-only Advertised auto-negotiation: Yes Link partner advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Link partner advertised pause frame use: Symmetric Receive-only Link partner advertised auto-negotiation: Yes Speed: 1000Mb/s Duplex: Full Port: MII PHYAD: 0 Transceiver: external Auto-negotiation: on Supports Wake-on: ug Wake-on: g Link detected: yes
$ sudo ethtool -i eth0 driver: eqos version: firmware-version: expansion-rom-version: bus-info: 2490000.ether_qos supports-statistics: yes supports-test: no supports-eeprom-access: no supports-register-dump: no supports-priv-flags: no
Thanks. I was needing to find the register definitions for the Ethernet controller so I could write a driver for our OS. I could not find it in the Parker TRM available online so I assumed it was an external chip from the SoC. Anyway, I was pointed to an older TRM that was obtained under NDA from a co-worker and it has the Ethernet controller documentation I needed.
I received my TX2 Dev kit today, am reading through the quick start guide PDF. On page 4 the guide says:
• A Jetson TX2 Developer Kit
• Your Jetson TX2 carrier board must be cabled as follows:
• Serial cable plugged into the serial port on the target connected to your Linux host directly…
Am I correct that I need a a separate Linux host running on x86_64 to develop for the TX2 Dev kit which already runs Ubuntu on aarch64?
Tools for flashing and upgrading require any x86_64 PC host. If you use the JetPack GUI tool for those operations (which is recommended), then it needs to be an Ubuntu 14.04 or 16.04 host.
Many of the things a developer would want to do are far more convenient if you have a Linux host as well. For example, the newest of the L4T releases allow you to compile the kernel natively on the Jetson, but some of the prior releases would have required you to either have a PC host or go through a painful process of adding foreign architecture support to the Jetson…once set up the PC host is easy to work with for kernel builds.
If you are building CUDA applications you will also find the host more convenient (naming the aarch64 as a target). Lots of IDEs are available on the host, but embedded systems are more limited, so something like eclipse on the Jetson would be difficult to use. The “nsight” edition of eclipse which supports CUDA is ported only to x86_64.
If you just want to do normal user space development and do not want to upgrade the Jetson’s o/s, then you can get away with no PC…I will discourage that since the shipped L4T is much less reliable than the newer versions.
Can you provide an eMMC schematic in the module?
I want to analyze eMMC behavior.