Single-Event Effects on Commercial-Off-the-Shelf Edge-Processing Artificial Intelligence ASICs

Megan C. Casey, Justin Goodwill, Edward J. Wyrwas, Rebekah Austin, Christopher M. Wilson, Scott D. Stansberry, Nicolas Gorius, Shahid Aslam

2023IEEE Transactions on Nuclear Science18 citationsDOI

Abstract

Single-event effects (SEEs) test results are presented for heavy ions and high-energy protons on two commercial edge-processing artificial intelligence (AI) application-specific integrated circuits (ASICs), particularly the Google Coral Accelerator Module, which employs a Google Coral Edge Tensor Processing Unit (TPU), and the Intel Neural Compute Stick 2 (NCS2), which employs the Intel Movidius Myriad X. The observed SEE signatures, probable causes within the devices, and single-event functional interrupt (SEFI) cross sections are discussed. Ultimately, when single-event upsets (SEUs) or SEFIs are observed, there are negligible changes in the measured precision or accuracy of the algorithm, and rarely do the errors result in incorrect identification or classification.

Topics & Concepts

InterruptEvent (particle physics)Single event upsetComputer scienceApplication-specific integrated circuitEnhanced Data Rates for GSM EvolutionComplex event processingArtificial neural networkComputer hardwareEmbedded systemArtificial intelligencePhysicsOperating systemProcess (computing)MicrocontrollerStatic random-access memoryQuantum mechanicsRadiation Effects in ElectronicsRadiation Detection and Scintillator TechnologiesIntegrated Circuits and Semiconductor Failure Analysis