RangeGuard: Efficient, Bounded Approximate Error Correction for Reliable DNNs

TL;DR

RangeGuard is a metadata-based error correction framework that encodes numerical range information to efficiently detect and correct errors in DNNs, ensuring reliability with minimal redundancy.

Contribution

It introduces RangeGuard, a novel approach using Range Identifiers to provide bounded approximate correction for DNNs with high efficiency and low overhead.

Findings

RangeGuard tolerates 64+ bit flips with only 16 bits of parity.

It maintains DNN accuracy despite frequent memory errors.

RangeGuard effectively bounds error magnitudes within numerical ranges.

Abstract

As DRAM scales in density and adopts 3D integration, raw fault rates increase and multi-bit errors are no longer rare. Such errors can severely impact Deep Neural Networks (DNNs): although DNNs tolerate small numerical perturbations, random bit flips can create extreme outliers that propagate and sharply degrade accuracy. Large Language Models (LLMs) are particularly vulnerable because attention, residual, and normalization layers can amplify and preserve a single corrupted activation across many layers, destabilizing inference. This paper introduces RangeGuard, a metadata-centric error-correcting framework that provides strong reliability and high efficiency based on bounded approximate correction. Instead of protecting raw bits, RangeGuard encodes compact Range Identifiers (RIDs) that capture the numerical range of each value. These compact metadata enable efficient use of limited redundancy and concentrate protection on range changes, which indicate harmful semantic deviations, while ignoring benign intra-range variations. Upon detecting a range change, RangeGuard restores the correct range and substitutes a representative value, ensuring that error magnitudes are bounded within the range. Based on RIDs, RangeGuard can tolerate 64+ flipped bits using only 16 bits of parity available in GPU memories without a noticeable accuracy loss. By introducing semantic range protection, RangeGuard enables reliable DNN execution even under frequent memory errors and tight redundancy budgets.