The Deep Neural Network Node, or DNN-Node enables older network cameras to compute at the edge in tedious, dangerous and subjective visual tasks such as counting, sorting, and observational reporting. It is the same tech found in the DNNCam, just without the camera.
This powerful device is an ultra power efficient, supercomputing camera with special functions designed for extremely rugged environments. It contains all necessary elements to sample enormous high resolution scenes, reduce the data through deep learning (Caffe, Tensorflow, TensorRT, CuDNN), and store or transmit highly reduced, contextual information.
The GPU can leverage the massive Nvidia ecosystem- the network effect of several developers working to build upon the latest tech. Take advantage of a decade of work
Each and every DNN-Node is vacuum tested down to -1 ATM (15 PSI) to ensure no leaks in the enclosure. The system is rated to IP-67 (water splashes, dust). However, it can survive in much more harsh conditions. The 15-W power consumption is generally enough to keep the lens defogged in cold conditions.
The DNN-Node is designed to be passively cooled, using computational fluid dynamics simulation below and direct measurement, the GPU core inside the TX2 system shows a stable temperature of just 27C above ambient. According to the Nvidia Jetson TX2 Thermal Design Guide, the GPU core maximum temperature is 93C. This represents a theoretical maximum ambient temperature of 65C or 149F. Boulder AI publishes the maximum operating temp of the DNN-Node to be 60C (140F).
The DNN-Node is a modular system. Inside, the camera consists of a series of building blocks that enable it to be configurable in many ways.
The price listed is the unit price to 10 units, volume budgetary pricing above 10 units is as follows:
10+ units $1499 /ea
100+ units $1299 /ea