OpenACMv2: An Accuracy-Constrained Co-Optimization Framework for Approximate DCiM

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhou, Yiqi, Yuan, Yue, Wang, Yikai, Liu, Bohao, Mei, Qinxin, Liu, Zhuohua, Shen, Shan, Xing, Wei, Sun, Daying, Li, Li, Liu, Guozhu
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914391223435264
author Zhou, Yiqi
Yuan, Yue
Wang, Yikai
Liu, Bohao
Mei, Qinxin
Liu, Zhuohua
Shen, Shan
Xing, Wei
Sun, Daying
Li, Li
Liu, Guozhu
author_facet Zhou, Yiqi
Yuan, Yue
Wang, Yikai
Liu, Bohao
Mei, Qinxin
Liu, Zhuohua
Shen, Shan
Xing, Wei
Sun, Daying
Li, Li
Liu, Guozhu
contents Digital Compute-in-Memory (DCiM) accelerates neural networks by reducing data movement. Approximate DCiM can further improve power-performance-area (PPA), but demands accuracy-constrained co-optimization across coupled architecture and transistor-level choices. Building on OpenYield, we introduce Accuracy-Constrained Co-Optimization (ACCO) and present OpenACMv2, an open framework that operationalizes ACCO via two-level optimization: (1) accuracy-constrained architecture search of compressor combinations and SRAM macro parameters, driven by a fast GNN-based surrogate for PPA and error; and (2) variation- and PVT-aware transistor sizing for standard cells and SRAM bitcells using Monte Carlo. By decoupling ACCO into architecture-level exploration and circuit-level sizing, OpenACMv2 integrates classic single- and multi-objective optimizers to deliver strong PPA-accuracy tradeoffs and robust convergence. The workflow is compatible with FreePDK45 and OpenROAD, supporting reproducible evaluation and easy adoption. Experiments demonstrate significant PPA improvements under controlled accuracy budgets, enabling rapid "what-if" exploration for approximate DCiM. The framework is available on https://github.com/ShenShan123/OpenACM.
format Preprint
id arxiv_https___arxiv_org_abs_2603_13042
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle OpenACMv2: An Accuracy-Constrained Co-Optimization Framework for Approximate DCiM
Zhou, Yiqi
Yuan, Yue
Wang, Yikai
Liu, Bohao
Mei, Qinxin
Liu, Zhuohua
Shen, Shan
Xing, Wei
Sun, Daying
Li, Li
Liu, Guozhu
Machine Learning
Hardware Architecture
Digital Compute-in-Memory (DCiM) accelerates neural networks by reducing data movement. Approximate DCiM can further improve power-performance-area (PPA), but demands accuracy-constrained co-optimization across coupled architecture and transistor-level choices. Building on OpenYield, we introduce Accuracy-Constrained Co-Optimization (ACCO) and present OpenACMv2, an open framework that operationalizes ACCO via two-level optimization: (1) accuracy-constrained architecture search of compressor combinations and SRAM macro parameters, driven by a fast GNN-based surrogate for PPA and error; and (2) variation- and PVT-aware transistor sizing for standard cells and SRAM bitcells using Monte Carlo. By decoupling ACCO into architecture-level exploration and circuit-level sizing, OpenACMv2 integrates classic single- and multi-objective optimizers to deliver strong PPA-accuracy tradeoffs and robust convergence. The workflow is compatible with FreePDK45 and OpenROAD, supporting reproducible evaluation and easy adoption. Experiments demonstrate significant PPA improvements under controlled accuracy budgets, enabling rapid "what-if" exploration for approximate DCiM. The framework is available on https://github.com/ShenShan123/OpenACM.
title OpenACMv2: An Accuracy-Constrained Co-Optimization Framework for Approximate DCiM
topic Machine Learning
Hardware Architecture
url https://arxiv.org/abs/2603.13042