Nvidia Jetson Xavier NX vs TX2 NX

In this article, we’re going to review Nvidia Jetson’s two latest Single Board Computers (SBCs): Nvidia Jetson Xavier NX and Nvidia Jetson TX2 NX.

You might think that the Jetson TX2 isn’t even comparable to the Xavier NX, and you’d be right… which is why it’s important to understand that there are two TX2 versions: the original Jetson TX2 (which launched a few years ago) and the brand new TX2 NX. We’ll be discussing the merits of the latter.

Let’s dive in to see what they’re all about.

NVIDIA Jetson Xavier NX

NVIDIA Jetson Xavier NX was launched in 2020 as part of NVIDIA’S IoT single-board lines. The company describes it “as the world’s smallest supercomputer” and is available with a Developer Kit that provides the core system-on-system (SOM) board with an expansion baseboard.

From a performance perspective, the developers kit is able to run AI workloads that deliver “modern neural networks in parallel and process data from multiple high-resolution sensors—a requirement for full AI systems.” Source: Nvidia.com

Combined with the developer kit you get access to NVIDIA’S “Cloud Native Computing”. The vision is to turn machine learning developers on to separating applications from operating systems. This allows either to be updated at will and new applications deployed quickly.

Here’s a list of the features of NVIDIA Jetson Xavier NX Module:

70 mm x 45 mm module size
AI Performance: 10 W = 14 TOPS (INT8) 15 W = 21 TOPS (INT8)
6-core NVIDIA Carmel ARM®v8.2 64-bit CPU 6MB L2 + 4MB L3
8GB LPDDR4x 128-bit RAM @ 1600 MHz offering 51.2 gigabytes per second
384-core NVIDIA Volta™ GPU with 48 Tensor Cores
16GB eMMC 5.1 Flash Storage

As well as all of these features, the NVIDIA Jetson Xavier NX comes with multiple neural networks which run in parallel and process data from multiple high-resolution sensors, perfect for embedded and edge computing devices. It consumes as little as 10 watts of power, too.

NVIDIA Jetson TX2 NX

NVIDIA Jetson TX2 NX was launched in 2021 as part of NVIDIA’S IoT single-board lines and is a 260-pin variant on the Jetson TX2. As the company puts it, it “delivers the next step in AI performance for entry-level embedded and edge products.” It uses the same Developer Kit that is available with the Xavier NX.

From a performance perspective, the module “provides up to 2.5X the performance of Jetson Nano, and shares form-factor and pin compatibility with Jetson Nano and Jetson Xavier™ NX.” Source: Nvidia.com

The hardware packs a punch for such a small module. Fully customizable, AI network development deployment is easy, along with flexible and seamless centralized updates.

Here’s a list of the features of NVIDIA Jetson TX2 NX Module:

69.6 mm x 45 mm module size
AI Performance 1.33 TFLOPs
Dual-core NVIDIA Denver 2 64-bit CPU and quad-core ARM A57 Complex
4GB LPDDR4 128-bit RAM 1600 MHz offering 51.2 gigabytes per second
NVIDIA Pascal™ Architecture GPU with 256 CUDA cores
16GB eMMC 5.1 Flash Storage

As well as all of these features, the NVIDIA Jetson TX2 NX is designed to be used with AI development and learning so the SBC supports NVIDIA’s JetPack SDK, Cafe, and Tensorflow.

NVIDIA Jetson Xavier NX vs NVIDIA Jetson TX2 NX

While these SBCs features are similar; they are both compact single-board computers, there are some subtle differences.

The TX2 NX features an ARM processor as does the Xavier NX. On the other hand, the Xavier NX has 8GB of memory while the TX2 NX only has 4GB.

The TX2 NX is designed to sit in between the Jetson Nano and Jetson TX2.

In terms of the GPU, the Xavier NX wins because of its 384-core NVIDIA Volta™ GPU with 48 Tensor Cores. The TX2 NX is weaker with only a 256-core NVIDIA Pascal™ Architecture GPU.

Comparing the CPUs, the Xavier NX uses the latest and best CPU, the 6-core NVIDIA Carmel ARM®v8.2 64-bit CPU. Although the TX2 NX does use a Quad-core ARM A57 Complex CPU, the Xavier NX CPU is superior.

And while both CPUs provide high performance and fast clocking speeds, it does depend on what solution you’re looking for.

The Xavier NX is great for embedded and edge computing devices that demand increased performance but are constrained by size, weight, and power budgets. Whereas the TX2 NX provides up to 2.5X the performance of Jetson Nano and shares form-factor and pin compatibility with Jetson Nano and Jetson Xavier™ NX.

When looking at price comparison, the Jetson TX2 NX is available for $199 whereas the Xavier NX costs $479, and there can be a long delay in delivery of the Xavier NX.

The Xavier NX is a more robust SBC but not every business or application requires the features that the Xavier NX provides, especially at the cost. Both are compatible with the Xavier NX Developer Kit, which is an additional $399.

Final Thoughts

While both of these SBCs are comparable, if you need higher GPU power and processing or you need a high-quality industrial solution, the NVIDIA Xavier NX is the best option to use.

If you don’t need the higher GPU power, then the TX2 NX is a great solution, especially if you’re looking to upgrade your Jetson Nano. You’ll get instant access to 2.5X the performance but retain the small size and save yourself some dollars.

And if you combine the TX2 NX with the Xavier NX Developer Kit, you’re getting access to more memory and processing power at a fraction of the cost. It provides 10X the performance of its widely adopted predecessor Jetson TX2.

If you want to utilize NVIDIA’s new cloud-native support, the developer kit will provide this capability to both the Xavier NX and TX2 NX.

No matter what your requirements are when you join JFrog Connect today, you’ll be able to easily and securely manage your edge products within minutes. Get in touch with us to discuss your next steps.