| Technical Description | Low-power deep learning accelerator integrates neural network design / training, model compressionaccelerator design. It uses a small amount of computationmemory footprint to realize high performance on the edge device.The image semantic segmentation technology can reach 80 frames per second (resolution:1024*2048) to meet real-time requirements for autonomous driving. | ||
|---|---|---|---|
| Scientific Breakthrough | 1. An AI engine that accelerates multi-layer fusion neural networks (each layer can be DNN, RNN, GRULSTM)2. Parameterized number of neurons for input/hidden/output layers 3. Smart model compression achieving 2 to 16 times compression ratio4. Configurable decimal point position for input, weight/biasoutput of each layer5. 8 to 256 Configurable MACs6. Easy to use SDK | ||
| Industrial Applicability | Low-power deep learning accelerator can be widely used in IC design, communications, transportation, home appliances, consumer electronics, e-health etc. related industry. The image semantic segmentationobject detection technology can be used in autonomous driving, medical diagnosis, security surveillance, human-computer-interface, etc. | ||