← 返回 JSSC 论文列表JSSC 2024第7期Memory65nmCIM
PRESTO: A Processing-in-Memory-Based k-SAT Solver Using Recurrent Stochastic Neural Network With Unsupervised Learning
基于存内计算的k-SAT求解器PRESTO，采用随机神经网络和混合信号电路架构
65nm CMOS, 0.4mm²核心面积, 100-500MHz工作频率, 35.4mW峰值功耗
存内计算随机神经网络k-SAT求解器混合信号电路无监督学习
▸创新点1：基于存内计算的混合信号电路架构（系统创新）。该设计采用混合信号电路存内计算（MSC-PIM）架构，将模拟计算与数字计算结合，显著提升了计算效率和能效比，核心面积仅为0.4 mm²，峰值功耗为35.4 mW。
▸创新点2：支持混合k-SAT问题的随机神经网络（方法创新）。PRESTO利用随机神经网络处理混合k-SAT问题，通过完全连接的k-SAT子句支持，展现了其在处理复杂SAT问题上的广泛适用性和高效性。
▸创新点3：无监督学习的硬件实现（电路创新）。该设计在硬件层面实现了无监督学习机制，通过随机神经网络的动态调整，无需外部监督即可优化求解过程，提升了系统的自主性和适应性。
▸创新点4：高性能与低功耗的平衡（系统创新）。PRESTO在65-nm CMOS工艺下实现，工作频率范围为100-500 MHz，在三-SAT问题上达到74.0%的准确率，展现了高性能与低功耗的优异平衡。
Abstract
In this article, we introduce a processing-in-memory (PIM)-based satisfiability (SAT) solver [called Processing-in- memory-based SAT solver using a Recurrent Stochastic neural network (PRESTO)], a mixed-signal circuit-based PIM (MSC- PIM) architecture combined with a digital finite state machine (FSM) for solving SAT problems. The presented design leverages a stochastic neural network with unsupervised learning. PRESTO’s architecture supports fully connected k-SAT clauses with mixed- k problems, highlighting its versatility in handling a wide range of SAT challenges. A test chip is fabricated in 65-nm CMOS technology with a core size of 0.4 mm 2 and demonstrates an operating frequency range of 100–500 MHz and a peak power of 35.4 mW. The measurement results show that PRESTO achieves a 74.0% accuracy for three-SAT problems with 30 variables and 126 clauses.