The Quest for Reliable AI Accelerators: Cross-Layer Evaluation and Design Optimization
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909995863375872 |
|---|---|
| author | Li, Meng Xie, Tong Zhang, Zuodong Wang, Runsheng |
| author_facet | Li, Meng Xie, Tong Zhang, Zuodong Wang, Runsheng |
| contents | As the CMOS technology pushes to the nanoscale, aging effects and process variations have become increasingly pronounced, posing significant reliability challenges for AI accelerators. Traditional guardband-based design approaches, which rely on pessimistic timing margin, sacrifice significant performance and computational efficiency, rendering them inadequate for high-performance AI computing demands. Current reliability-aware AI accelerator design faces two core challenges: (1) the lack of systematic cross-layer analysis tools to capture coupling reliability effects across device, circuit, architecture, and application layers; and (2) the fundamental trade-off between conventional reliability optimization and computational efficiency. To address these challenges, this paper systematically presents a series of reliability-aware accelerator designs, encompassing (1) aging and variation-aware dynamic timing analyzer, (2) accelerator dataflow optimization using critical input pattern reduction, and (3) resilience characterization and novel architecture design for large language models (LLMs). By tightly integrating cross-layer reliability modeling and AI workload characteristics, these co-optimization approaches effectively achieve reliable and efficient AI acceleration. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_14148 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | The Quest for Reliable AI Accelerators: Cross-Layer Evaluation and Design Optimization Li, Meng Xie, Tong Zhang, Zuodong Wang, Runsheng Hardware Architecture As the CMOS technology pushes to the nanoscale, aging effects and process variations have become increasingly pronounced, posing significant reliability challenges for AI accelerators. Traditional guardband-based design approaches, which rely on pessimistic timing margin, sacrifice significant performance and computational efficiency, rendering them inadequate for high-performance AI computing demands. Current reliability-aware AI accelerator design faces two core challenges: (1) the lack of systematic cross-layer analysis tools to capture coupling reliability effects across device, circuit, architecture, and application layers; and (2) the fundamental trade-off between conventional reliability optimization and computational efficiency. To address these challenges, this paper systematically presents a series of reliability-aware accelerator designs, encompassing (1) aging and variation-aware dynamic timing analyzer, (2) accelerator dataflow optimization using critical input pattern reduction, and (3) resilience characterization and novel architecture design for large language models (LLMs). By tightly integrating cross-layer reliability modeling and AI workload characteristics, these co-optimization approaches effectively achieve reliable and efficient AI acceleration. |
| title | The Quest for Reliable AI Accelerators: Cross-Layer Evaluation and Design Optimization |
| topic | Hardware Architecture |
| url | https://arxiv.org/abs/2601.14148 |