FSL-HDnn: A 40 nm Few-shot On-Device Learning Accelerator with Integrated Feature Extraction and Hyperdimensional Computing

TL;DR

FSL-HDnn is a 40 nm energy-efficient accelerator designed for on-device few-shot learning, combining feature extraction and hyperdimensional computing to enable fast, low-energy training and inference on resource-limited edge devices.

Contribution

The paper presents a novel integrated accelerator that combines feature extraction with hyperdimensional computing for efficient on-device few-shot learning, reducing latency and energy consumption.

Findings

Achieves 6 mJ/image training energy efficiency

Provides 28 images/sec throughput for 10-way 5-shot tasks

Reduces training latency by up to 20 times compared to existing chips

Abstract

This paper introduces FSL-HDnn, an energy-efficient accelerator that implements the end-to-end pipeline of feature extraction and on-device few-shot learning (FSL). The accelerator addresses fundamental challenges of on-device learning (ODL) for resource-constrained edge applications through two synergistic modules: a parameter-efficient feature extractor employing weight clustering and an FSL classifier based on hyperdimensional computing (HDC). The feature extractor exploits the weight clustering mechanism to reduce computational complexity, while the HDC-based FSL classifier eliminates gradient-based back propagation operations, enabling single-pass training with substantially reduced latency. Additionally, FSL-HDnn enables low-latency ODL and inference via two proposed optimization strategies, including an early-exit mechanism with branch feature extraction and batched single-pass training that improves hardware utilization. Measurement results demonstrate that our chip fabricated in a 40 nm CMOS process delivers superior training energy efficiency of 6 mJ/image and end-to-end training throughput of 28 images/s on a 10-way 5-shot FSL task. The end-to-end training latency is also reduced by 2x to 20.9x compared to state-of-the-art ODL chips.