SparkXD: A Framework for Resilient and Energy-Efficient Spiking Neural Network Inference using Approximate DRAM

TL;DR

SparkXD is a framework that enhances energy efficiency in Spiking Neural Network inference by using approximate DRAM with error mitigation techniques, achieving around 40% DRAM energy reduction while maintaining accuracy.

Contribution

It introduces a comprehensive approach combining fault-aware training, error tolerance analysis, and optimized data mapping to enable resilient, energy-efficient SNN inference with approximate DRAM.

Findings

Achieves approximately 40% reduction in DRAM energy consumption.

Maintains within 1% of baseline accuracy despite DRAM errors.

Provides a systematic framework for resilient SNN inference using approximate memory.

Abstract

Spiking Neural Networks (SNNs) have the potential for achieving low energy consumption due to their biologically sparse computation. Several studies have shown that the off-chip memory (DRAM) accesses are the most energy-consuming operations in SNN processing. However, state-of-the-art in SNN systems do not optimize the DRAM energy-per-access, thereby hindering achieving high energy-efficiency. To substantially minimize the DRAM energy-per-access, a key knob is to reduce the DRAM supply voltage but this may lead to DRAM errors (i.e., the so-called approximate DRAM). Towards this, we propose SparkXD, a novel framework that provides a comprehensive conjoint solution for resilient and energy-efficient SNN inference using low-power DRAMs subjected to voltage-induced errors. The key mechanisms of SparkXD are: (1) improving the SNN error tolerance through fault-aware training that considers bit errors from approximate DRAM, (2) analyzing the error tolerance of the improved SNN model to find the maximum tolerable bit error rate (BER) that meets the targeted accuracy constraint, and (3) energy-efficient DRAM data mapping for the resilient SNN model that maps the weights in the appropriate DRAM location to minimize the DRAM access energy. Through these mechanisms, SparkXD mitigates the negative impact of DRAM (approximation) errors, and provides the required accuracy. The experimental results show that, for a target accuracy within 1% of the baseline design (i.e., SNN without DRAM errors), SparkXD reduces the DRAM energy by ca. 40% on average across different network sizes.