Ising-CIM: A Reconfigurable and Scalable Compute Within Memory Analog Ising Accelerator for Solving Combinatorial Optimization Problems

Ising-CIM: A Reconfigurable and Scalable Compute Within Memory Analog Ising Accelerator for Solving Combinatorial Optimization Problems
复制标题

Ising-CIM:用于解决组合优化问题的可重新配置和可扩展的内存模拟 Ising 加速器

DOI:
10.1109/jssc.2022.3176610
复制
发表时间:
2022
影响因子:
5.4
通讯作者:
Kulkarni, Jaydeep P.
Kulkarni, Jaydeep P.
中科院分区:
工程技术1区
文献类型:
--
作者:
Xie, Shanshan;Raman, Siddhartha Raman;Ni, Can;Wang, Meizhi;Yang, Mengtian;Kulkarni, Jaydeep P.

文献摘要

参考文献

被引文献

相似文献

组合优化问题(COPs)在现实世界的科学,工业和社会场景中找到应用。这样的COP在计算上是NP难的,并且随着COP大小的增加,执行穷举蛮力搜索以获得最优解变得站不住脚。为了加快COP的计算,伊辛模型的形式主义,抽象的自旋动力学的铁磁体。自旋被定向以达到最小能量状态,代表最佳COP解决方案。以前的伊辛引擎设计利用专用退火处理器或存储器位单元旁边的附加数字算术电路。这些定制电路或处理器无法重新用于其他应用,从而导致大量面积和功耗开销。与现有的方法相比,这项工作提出了一种可重构和可扩展的计算内存模拟方法的伊辛计算(称为伊辛CIM)。这种面积高效的方法重新利用现有的嵌入式存储器位单元列和外围电路来对位线执行模拟域哈密顿计算,从而显著地最小化面积和功率开销。一个13.18 Kb的硅原型,在65纳米CMOS工艺实现,演示了伊辛CIM的概念和功能,使用10064像素的图像在最大切割COP。与CPU相比,Ising-CIM设计实现了48-/自旋单位自旋面积和加速退火时间。
Combinatorial optimization problems (COPs) find applications in real-world scientific, industrial, and societal scenarios. Such COPs are computationally NP-hard, and performing an exhaustive brute force search for the optimal solution becomes untenable as the COP size increases. To expedite the COP computation, the Ising model formalism is used, which abstracts spin dynamics in a ferromagnet. The spins are orientated to reach the minimum energy state, representing the optimum COP solution. Previous Ising engine designs utilized dedicated annealing processors or additional digital arithmetic circuits next to the memory bitcells. These custom circuits or processors cannot be repurposed for other applications, incurring significant area and power overhead. In contrast to the prior approaches, this work presents a reconfigurable and scalable compute-within-memory analog approach for Ising computation (called Ising-CIM). This area-efficient approach repurposes existing embedded memory bitcell columns and peripheral circuits to perform analog domain Hamiltonian calculations on the bitlines minimizing area and power overhead significantly. A 13.18-Kb silicon prototype, implemented in a 65-nm CMOS process, demonstrates the Ising-CIM concept and functionality using a 10064 pixel image in a max-cut COP. The Ising-CIM design achieves 48-/spin unit spin area andspeedup in annealing time compared to the CPU.
2.6 基于内存处理方法的 2 ×30k 旋转多芯片可扩展退火处理器,用于解决大规模组合优化问题
DOI: --
发表时间: 2019
期刊: IEEE International Solid-State Circuits Conference
影响因子: --
作者:
Takashi Takemoto;Masato Hayashi;C. Yoshimura;M. Yamaoka
通讯作者: M. Yamaoka
PHLOGON:使用振荡纳米系统的基于相位的逻辑
DOI: 10.1007/978-3-319-08123-6_29
发表时间: 2014
期刊: IEICE Trans. Inf. Syst.
影响因子: --
作者:
Tianshi Wang;J. Roychowdhury
通讯作者: J. Roychowdhury
纳米级 SRAM 中技术与设计的相互作用
DOI: --
发表时间: 2009
期刊:
影响因子: --
作者:
R. Mann
通讯作者: R. Mann
DOI: 10.1109/jxcdc.2020.3046558
发表时间: 2020-12-01
影响因子: 2.4
作者:
Csaba, Gyorgy;Porod, Wolfgang
通讯作者: Porod, Wolfgang
4.6 由 9 个 16Kb 退火处理器芯片组成的 144Kb 退火系统,具有可扩展的芯片到芯片连接,用于解决大规模组合优化问题
DOI: --
发表时间: 2021
期刊: IEEE International Solid-State Circuits Conference
影响因子: --
作者:
Takashi Takemoto;Kasho Yamamoto;C. Yoshimura;Masato Hayashi;Masafumi Tada;Hiroaki Saito;Mayumi Mashimo;M. Yamaoka
通讯作者: M. Yamaoka