ACCFormer: Predicting Analog Circuit Performance Metrics via Topology-Aware Transformers

Liao, Bowen; Feng, Yutong; Lin, Jianhua; Wu, Zhaohui; Liang, Yuxuan; Li, Bin

doi:10.24963/ijcai.2026/720

ACCFormer: Predicting Analog Circuit Performance Metrics via Topology-Aware Transformers

Bowen Liao, Yutong Feng, Jianhua Lin, Zhaohui Wu, Yuxuan Liang, Bin Li

Download BibTeX Cite

Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence. AI4Tech: AI Enabling Technologies. Pages 6471-6479.

https://doi.org/10.24963/ijcai.2026/720

Keywords

Advanced AI4Tech: Data-driven AI4Tech
Domain-specific AI4Tech: Other AI4Tech applications
AI4Tech infrastructure/systems: AI chips, AI sensors, AI computers

Abstract

Reusing and migrating analog circuit intellectual property (IP) across process nodes poses a significant challenge in modern chip design. Efficient and generalizable circuit performance prediction methods for analog circuits are crucial to achieving this goal. Current data-driven approaches typically rely on manually designed features, which perform poorly on unseen circuit architectures and struggle to model the inherent structural relationships within analog designs. To address these challenges, we propose ACCFormer, a novel topology-aware Transformer framework for predicting performance metrics of analog circuit. Our model combines device parameters with connectivity data to learn topology-aware representations, followed by a performance-oriented cross-attention mechanism where trainable metric queries adaptively focus on the most critical devices for each target parameter. Validated across different process nodes, our model achieves state-of-the-art prediction accuracy and demonstrates strong cross-process adaptability, highlighting its potential to accelerate IP reuse and reduce design cycles.

Bowen Liao, Yutong Feng, Jianhua Lin, Zhaohui Wu, Yuxuan Liang, and Bin Li. 2026. ACCFormer: Predicting Analog Circuit Performance Metrics via Topology-Aware Transformers. In Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26. IJCAI, 6471–6479. https://doi.org/10.24963/ijcai.2026/720

@inproceedings{ijcai2026p720,
  title     = {ACCFormer: Predicting Analog Circuit Performance Metrics via Topology-Aware Transformers},
  author    = {Liao, Bowen and Feng, Yutong and Lin, Jianhua and Wu, Zhaohui and Liang, Yuxuan and Li, Bin},
  booktitle = {Proceedings of the Thirty-Fifth International Joint Conference on
               Artificial Intelligence, {IJCAI-26}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Diego Calvanese},
  pages     = {6471--6479},
  year      = {2026},
  month     = {8},
  note      = {AI4Tech: AI Enabling Technologies},
  doi       = {10.24963/ijcai.2026/720},
  url       = {https://doi.org/10.24963/ijcai.2026/720},
}

TY  - CPAPER
TI  - ACCFormer: Predicting Analog Circuit Performance Metrics via Topology-Aware Transformers
AU  - Liao, Bowen
AU  - Feng, Yutong
AU  - Lin, Jianhua
AU  - Wu, Zhaohui
AU  - Liang, Yuxuan
AU  - Li, Bin
T2  - Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2026
SP  - 6471
EP  - 6479
DO  - 10.24963/ijcai.2026/720
UR  - https://doi.org/10.24963/ijcai.2026/720
ER  -