Nvidia has added the compact new Jetson Xavier NX to its Jetson product family. In essence this is a miniature compute-module product, smaller than a credit card at 70 x 45mm, but it boasts "server-class performance up to 21 TOPS for running modern AI workloads, and consumes as little as 10 watts of power". This makes it the "world's smallest supercomputer for AI at the edge," boasts Nvidia.
It is intended that the Jetson Xavier NX boards will find themselves in embedded edge computing devices that would benefit from the increased performance but would have previously eschewed better processing power due to size and/or weight and/or power constraints (or cost). Nvidia foresees a host of small commercial robots, drones, intelligent factory sensing and monitoring equipment, portable medical devices, and of industrial IoT devices taking up Jetson Xavier NX.
Nvidia's VP and GM of Edge Computing, Deepu Talla, asserted that "AI has become the enabling technology for modern robotics and embedded devices that will transform industries." Talla went on to claim that the new Jetson Xavier NX will be a key tool for this transformation as it allows partners to dramatically increase AI capabilities within their usual restraints.
Jetson Xavier NX runs on the same CUDA-X AI software architecture as all Jetson offerings. It is supported by the Nvidia JetPack software development kit which includes a complete AI software stack, accelerated libraries for deep learning as well as computer vision, computer graphics, multimedia and more. This compute module is configurable to deliver up to 14 TOPS (at 10W) or 21 TOPS (at 15W) - enough to running multiple neural networks in parallel and processing data from multiple high-resolution sensors simultaneously.
Jetson Xavier NX module specifications:
- GPU: Nvidia Volta with 384 CUDA cores and 48 Tensor Cores, plus 2x NVDLA
- CPU: 6-core Nvidia Carmel Arm 64-bit CPU, 6MB L2 + 4MB L3
- Display: 2x multi-mode DP 1.4/eDP 1.4/HDMI 2.0
- Video: 2x 4K30 Encode and 2x 4K60 Decode
- Camera: Up to six CSI cameras (36 via virtual channels); 12 lanes (3x4 or 6x2) MIPI CSI-2
- Memory: 8GB 128-bit LPDDR4x; 51.2GB/second
- Storage: 16GB eMMC 5.1
- Connectivity: Gigabit Ethernet
- OS Support: Ubuntu-based Linux
- Compatibility: pin-compatible with Jetson Nano as well as many peripherals and sensors
- AI software support: TensorFlow, PyTorch, MxNet, Caffe and others
- Module Size: 70x45mm
Nvidia will be making the Jetson Xavier NX widely available from March 2020, priced at US$399. Developers can begin application development immediately using the Jetson AGX Xavier Developer Kit with a software patch to emulate Jetson Xavier NX.