EdgeDRNN: Recurrent Neural Network Accelerator for Edge Inference

Abstract

Low-latency, low-power portable recurrent neural network (RNN) accelerators offer powerful inference capabilities for real-time applications such as IoT, robotics, and human-machine interaction. We propose a lightweight Gated Recurrent Unit (GRU)-based RNN accelerator called EdgeDRNN that is optimized for low-latency edge RNN inference with batch size of 1. EdgeDRNN adopts the spiking neural network inspired delta network algorithm to exploit temporal sparsity in RNNs. Weights are stored in inexpensive DRAM which enables EdgeDRNN to compute large multi-layer RNNs on the most inexpensive FPGA. The sparse updates reduce DRAM weight memory access by a factor of up to 10x and the delta can be varied dynamically to trade-off between latency and accuracy. EdgeDRNN updates a 5 million parameter 2-layer GRU-RNN in about 0.5ms. It achieves latency comparable with a 92W Nvidia 1080 GPU. It outperforms NVIDIA Jetson Nano, Jetson TX2 and Intel Neural Compute Stick 2 in latency by 5X. For a batch size of 1, EdgeDRNN achieves a mean effective throughput of 20.2GOp/s and a wall plug power efficiency that is over 4X higher than the commercial edge AI platforms.

References in corpus (6)

• Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Sequence Transduction with Recurrent Neural Networks
• Delta Networks for Optimized Recurrent Network Computation
• EdgeDRNN: Enabling Low-latency Recurrent Neural Network Edge Inference
• Recurrent Neural Network Control of a Hybrid Dynamic Transfemoral Prosthesis with EdgeDRNN Accelerator

Cited by in corpus (7)

• Empowering Edge Intelligence: A Comprehensive Survey on On-Device AI Models
• Automated Pest Detection with DNN on the Edge for Precision Agriculture
• Spartus: A 9.4 TOp/s FPGA-based LSTM Accelerator Exploiting Spatio-Temporal Sparsity
• MP-DPD: Low-Complexity Mixed-Precision Neural Networks for Energy-Efficient Digital Predistortion of Wideband Power Amplifiers
• DeltaKWS: A 65nm 36nJ/Decision Bio-inspired Temporal-Sparsity-Aware Digital Keyword Spotting IC with 0.6V Near-Threshold SRAM
• Spiking Cochlea with System-level Local Automatic Gain Control
• DPD-NeuralEngine: A 22-nm 6.6-TOPS/W/mm${}^2$ Recurrent Neural Network Accelerator for Wideband Power Amplifier Digital Pre-Distortion