NVIDIA® has Officially Launched the Jetson Xavier™ NX Development Kit. Here is What You Need to Know Before You Begin Your Project

May 26, 2020 Rob Callaghan, Director of Technical Services, Connect Tech Inc.

Until the release of the Jetson Xavier NX Development Kit, teams looking to determine if the module was a good fit for their project were required to use the AGX Xavier Development Kit with a patch. With both the production module and development kit commercially available for purchase, full-scale development using the Jetson Xavier NX module can be achieved. Here are the main differences between the production module and development kit for consideration:


NVIDIA’s Jetson Xavier NX development kit comes supplied with the entire NVIDIA software stack and development tools to accompany the module including a reference carrier board, AC power supply, and active heat sink. This kit allows for developers to build and test programs utilizing the newest module in the Jetson lineup. When teams are ready to bring in production modules for testing, they will quickly realize one key difference between the dev kit and production module: there is no thermal solution on the Jetson Xavier NX production module.

NVIDIA has changed the way that thermal relief is applied for the Jetson Nano™ and Jetson Xavier NX. With other products, like the Jetson™ TX2 and Jetson AGX Xavier™, the production module come equipped with an integrated thermal transfer plate (TTP). This allows teams to plan for further cooling beyond the TTP. Unique to the Jetson Nano and Jetson Xavier NX, NVIDIA does not include a TTP with production modules. Teams will need to either design their own thermal solution, or choose between available active heat sinks, passive heat sinks, or thermal transfer plates to accompany the module before moving to production. 


In the case of Jetson Xavier NX, the production module includes 16GB of on-board eMMC storage. When teams order a development kit, it is important to note that the module included in the kit does not include any storage. The supplied module can only be used by inserting an SD card to load the OS and applications. Since the production module includes 16GB of storage, the SD card should be the equivalent capacity or greater.

For projects requiring storage beyond the 16GB supplied with the Jetson Xavier NX production module, carrier board solutions for the module offer extra storage solutions including additional SD card or NVMe (M.2 M-Key) options. Other solutions such as SATA, PCIe, or MSATA can be worked into the carrier board for your project.  


At first glance, the pricing model for development kits versus production modules can cause some confusion. Put simply, NVIDIA has not designed development kits to be revenue generating - they’re a vehicle to drive module sales.

NVIDIA’s promoted price for the development kit and production module are the same, but it’s important to note that NVIDIA’s pricing for modules are at quantities of 1,000 or more. If looking to purchase through one of NVIDIA’s approved distribution partners, the cost for lower quantities of the production modules will be higher than NVIDIA promotes.

Similar to the AGX Xavier and TX2 Dev kits, there are a few differences as to why the production modules cost more than the developer kit when by all appearances the kit comes supplied with more peripherals. The first is the mentioned storage. Production modules have 16GB eMMC storage included on the board, which increases the cost. But the main contributor to the additional expense is the fact that the module supplied with development kits are not intended for use in a production environment. NVIDIA explains that the materials used in the development modules may not be of production quality. Each of these modules may have components change without notification and only basic validation in a constrained environment is performed. In contrast, the production modules are built for the full 5-year operating lifetime. It’s for these reasons that the suggested use of development kits be limited to early phases of development and not be used for in-field deployment. As projects progress, it becomes important to select a field-deployable carrier board and other assemblies that can withstand the environment to which they’re being deployed.

About the Author

Rob Callaghan is the Director of Technical Services at Connect Tech Inc. He has over 11 years of experience as a lead designer and project manager on numerous high-speed digital design and embedded products developed by Connect Tech. Rob received an Honors Bachelor of Engineering in Systems and Computing Engineering from the University of Guelph in Canada. He is an IEEE member, and a Licensed Professional Engineer.

Previous Article
How VPN Security Works and Why Use One for Safe Internet Connection
How VPN Security Works and Why Use One for Safe Internet Connection

Information security today is one of the main priorities within many organizations. This is especially true...

Next Article
COVID-19 Is Impacting the Embedded Computing Industry and its Supply Chain
COVID-19 Is Impacting the Embedded Computing Industry and its Supply Chain

We are now living in the craziest of times, certainly the craziest of my lifetime, and likely yours. Obviou...