Amorphica: 4-Replica 512 Fully Connected Spin 336MHz Metamorphic Annealer with Programmable Optimization Strategy and Compressed-Spin-Transfer Multi-Chip Extension

📋 论文概要

本文提出Amorphica，一种4副本512全连接自旋的336MHz可变形退火器，用于解决组合优化问题。通过可编程优化策略，系统能够灵活适应不同问题，提高求解效率。

💡 主要创新点

• 采用4副本全连接自旋架构，实现高并行度退火计算
• 引入可编程优化策略，支持多种退火算法动态切换

🏷 关键词

📄 原文摘要

Genta Inoue1, Akira Hyodo1, Ángel López García-Arias1, Kota Ando2, Bruno Hideki Fukushima-Kimura2, Ryota Yasudo3, Thiem Van Chu1, Masato Motomura1 Tokyo Institute of Technology, Yokohama, Japan, 2Hokkaido University, Sapporo, Japan, Kyoto University, Kyoto, Japan, *Equally Credited Authors (ECA) 1 3 Combinatorial optimization (CO) is vital for making wiser decisions and planning in our society. Annealing computation is a promising CO approach derived from an analogy to physical phenomena (Fig. 2.3.1). It represents a CO problem as an energy function, a quadratic form of {1, -1} vectors, where each binary element is called a (pseudo) spin. The spin vector is initialized randomly and is updated stochastically to find minimum energy states by gradually reducing the (pseudo) temperature. Local-connection annealers (quantum [1] and non-quantum [2-4]) have been constrained to spin models having only local inter-spin couplings. This restriction, however, severely limits their CO

👥 作者与机构

Kazushi Kawamura*1, Jaehoon Yu*1, Daiki Okonogi1, Satoru Jimbo1,