Paper details transistor aging impact on DNN accuracy

A new research paper details the impact of transistor aging on the accuracy of deep neural networks (DNNs) used in applications like image classification. Transistor aging slows down switching speeds, leading to timing violations and reduced inference accuracy. The paper proposes an aging-aware retraining methodology to create more resilient DNNs that maintain accuracy even with aggressive timing guardbands, and also touches on using short-term aging for hardware Trojan detection. AI

IMPACT This research could lead to more reliable AI hardware implementations, ensuring sustained performance over time.

RANK_REASON This is a research paper published on arXiv detailing a technical methodology. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• Deep Neural Networks
• Prashanth Krishnamurthy

• paper
• safety
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →