FLARE: One-Shot PE-Level Fault Localization in Systolic Arrays via Algebraic Test Vectors

TL;DR

This paper introduces a novel coprime test vector method for one-shot fault localization at the PE level in systolic arrays, significantly improving fault detection accuracy with minimal overhead.

Contribution

It proposes a purely algorithmic approach using coprime test vectors for precise PE-level fault localization without hardware redundancy.

Findings

Single test pass localizes faulty row with high probability for INT16 arrays up to 256x256.

A second pass with ratio computation achieves exact localization when needed.

For single-bit errors, odd coprime entries enable one-round exact localization.

Abstract

Systolic arrays are the dominant compute fabric for neural network inference. Prior work has addressed column-level fault detection efficiently with uniform test patterns, but row-level (PE-level) fault localization within a faulty column remains open without resorting to hardware redundancy. The fundamental obstacle is that uniform test inputs destroy per-row signatures: any test that activates every row equally cannot distinguish which row is the source of an observed deviation. In this paper, we propose a lightweight, purely algorithmic remedy based on coprime test vectors. By assigning pairwise coprime integers as test-input entries, a permanent weight-register fault produces a deviation whose divisibility signature uniquely identifies the faulty row. Under a general bounded error model, a single test pass localizes the faulty row with high probability. This error model covers a broader class of faults than what prior dataflow-aware testing work has primarily emphasized. When one round is insufficient, a second pass using a ratio computation achieves exact localization; for the special case of single-bit errors, odd coprime entries guarantee exact localization in one round. For INT16 arithmetic, a single test pass covers array sizes up to $256{\times}256$ with localization probability above $0.98$, at a test cost under $1\%$ of one inference GEMM tile.