A Current-Mode Implementation of A Nearest Neighbor STDP Synapse
A Current-Mode Implementation of A Nearest Neighbor STDP Synapse
复制标题
DOI：
10.1109/newcas57931.2023.10198113
复制
发表时间：
2023-06
期刊：
影响因子：
--
通讯作者：
A. Akwaboah;Ralph Etienne-Cummings
中科院分区：
文献类型：
--
作者：
A. Akwaboah;Ralph Etienne-Cummings
人工智能（AI）的破坏仍然有增无减，尽管在极端的计算要求。神经形态电路和系统为这种奢侈提供了灵丹妙药。为此，基于事件的学习，如尖峰神经网络（SNN）中的尖峰时间依赖可塑性（STDP）是一个活跃的研究领域。SNN中的Hebbian学习基本上涉及基于突触前和突触后神经活动之间的时间相关性的突触权重更新。虽然有两种实现STDP的方法，即全对全与最近邻（NN），但在生物相容性方面存在支持NN方法的强有力的论据。在本文中，我们提出了一种新的电流模式实现的突触后事件为基础的NN STDP为基础的突触。我们利用晶体管的亚阈值动态生成指数STDP的痕迹使用重新设计的对数域低通滤波器电路。突触权重运算包括加法和乘法，分别由基尔霍夫电流定律和跨导线性原理实现。模拟结果从NCSU台积电180纳米技术。最后，这里提出的想法对设计高效硬件以满足不断增长的人工智能训练和推理需求具有影响。
The Artificial Intelligence (AI) disruption continues unabated, albeit at extreme compute requirements. Neuromorphic circuits and systems offer a panacea for this extravagance. To this effect, event-based learning such as spike-timing-dependent plasticity (STDP) in spiking neural networks (SNNs) is an active area of research. Hebbian learning in SNNs fundamentally involves synaptic weight updates based on temporal correlations between pre-and post-synaptic neural activities. While there are broadly two approaches of realizing STDP, i.e. All-to-All versus Nearest Neighbor (NN), there exist strong arguments favoring the NN approach on the biologically plausibility front. In this paper, we present a novel current-mode implementation of a postsynaptic event-based NN STDP-based synapse. We leverage transistor subthreshold dynamics to generate exponential STDP traces using repurposed log-domain low-pass filter circuits. Synaptic weight operations involving addition and multiplications are achieved by the Kirchoff current law and the translinear principle respectively. Simulation results from the NCSU TSMC 180 nm technology are presented. Finally, the ideas presented here hold implications for engineering efficient hardware to meet the growing AI training and inference demands.