Gem5-Shark: Smaug-based Heterogeneous Deep Learning Accelerator Security and Robustness Evaluation and Characterization Framework

Summary

Gem5-Shark is a cycle-level simulation framework, built on Gem5-Aladdin's Smaug, for evaluating security threats to heterogeneous deep learning accelerators before deployment. It integrates eight threat vectors across dataset, compilation, SoC communication and accelerator execution stages, including two new compilation-stage attacks, Operator Attack (OPAtt) and Tiling Attack (TAtt). Across Minerva, LeNet-5, VGG-16 and pretrained LLMs up to 8B parameters, the attacks degrade accuracy by up to 97.7%, raise latency by up to 84× and implant backdoors with as few as 20 poisoned fine-tuning samples.

Mitigation

Two defenses are proposed: a compiler-stage tiling range check and a host-side finish-flag watchdog, each removing over 99% of the injected overhead on CNNs and LLMs.

Related items

• MediumTop MCP security resources — November 2026Similar attack · Adversa AI Blog
• MediumAI broke the cybersecurity world. What’s next?Similar attack · The Verge (AI)
• InfoCVE-2026-123357: GROBID command execution through crafted PDF filename in Linux conversionSimilar attack · NVD/CVE Database
• MediumCVE-2026-123851: JetBrains Junie GitHub Action prompt injection bypass via HTML encodingSimilar attack · NVD/CVE Database
• MediumShai-Hulud Descendant Surfaces in Malicious Ruby Gem That Steals Credentials and Poisons AI ToolsSimilar attack · Socket Blog