The Jetson TX1 and Jetson TX2 from NVIDIA put the power of an AI supercomputer in the palm of your hand. These development kits and boards have the power to enable more complex deep neural networks and offer support for the latest technology in graphics, GPU computing, computer vision and more. Here’s what they can do for you.
TX1 or TX2?
Let’s start with the TX1, which is plenty capable on its own—featuring NVIDIA Maxwell™ architecture, 256 NVIDIA CUDA® cores, 64-bit Quad ARM A57/2 MB L2 CPUs, and a power-efficient design. The module itself supports encoded 30 Hz 4K video and decoded 60 Hz 4k video with up to 6 cameras. It has 4 GB of 64-bit LPDDR4 memory on board and 16 GB of native flash storage. In short, that means the TX1 module includes and supports the latest technology for deep learning, computer vision, GPU computing, and graphics—making it ideal for embedded AI computing for all sorts of applications.
Even better, NVIDIA has already made an extremely robust carrier board and development kit that will help you start developing right away. The TX1 Development Kit comes flashed with a Linux environment, supports the complete NVIDIA development tool chain, and includes support for common APIs.
TX1 Dev Board Feature Stack
• An external SD card storage port
• Integrated antennas for quick and strong Wi-Fi connectivity
• An onboard 5 MP fixed focus MIPI CSI camera
• Display and camera expansion headers
• A huge forced air heat sink
• Useful connectors like M.2 Key E, SATA data and power, Ethernet, HDMI, USB Type A, micro USB, Power, and more.
As if that wasn’t exciting enough, the TX1 dev kit also comes with an AC adapter and power chord to get the board up and running fast, a USB micro-B to USB A cable, and a USB micro-B to female USB A cable. All of these features make it an ideal tool to develop AI applications ranging from facial recognition to robot training to enterprise computing solutions and more, and one that can fit most embedded development needs at a more economic price point.
The TX2: Next Level Development
The TX2 has all the capabilities of the already-powerful TX1, but with double the strength behind it and a few more additional features. Instead of Maxwell GPU architecture, the TX2 comes with NVIDIA’s Pascal™ architecture, which boasts 150 billion transistors built on bleeding-edge 16 nanometer FinFET fabrication technology. In NVIDIA’s own words, “Pascal is the most powerful computer architecture ever built inside a GPU.”
If your application needs to do some particularly heavy lifting, here’s what the TX2 can give you:
• An HMP Dual Denver 2/2 MB L2 CPU in addition to the Quad ARM A57/2 BM L2 CPU.
• Support for encoded 4K video at 60 Hz as opposed to 30 Hz.
• 8GB 128 bit memory at 59.7 GB/z—that’s twice the memory at twice the bits at twice the size.
• Camera serial interface lane support that is almost twice as fast.
• Better connectivity for peripheral devices through its serial expansion bus.
• Twice the native data storage (32 GB on board).
The TX2 Dev Kit offers the capabilities and support of the TX1 Dev Kit, but also supports the NVIDIA Jetpack SDK, which includes the BSP, libraries for deep learning, computer vision, GPU computing, multimedia processing, and more.
Ultimately, the needs of your application will determine which Jetson module or dev kit you choose. Rest assured that no matter which of these solutions you end up going with, you’ll have the power to bring science fiction into reality.
Update: Check out the new Jetson TX2i Module!
Click HERE to learn more about All Jetson Products