Energy-Efficient Moderate Precision Time-Domain Mixed-Signal Vector-by-Matrix Multiplier Exploiting 1T-1R Arrays

Abstract

The emerging mobile devices in the era of Internet-of-Things (IoT) require a dedicated processor to enable computationally intensive applications such as neuromorphic computing and signal processing. Vector-by-matrix multiplication is the most prominent operation in these applications. Therefore, there is a critical need for compact and ultralow-power vector-by-matrix multiplier (VMM) blocks to perform resource-intensive low-to-moderate precision computations. To this end, in this article, we propose a timedomain mixed-signal VMM exploiting a modified configuration of 1MOSFET-1RRAM (1T-1R) array. The proposed VMM overcomes the energy inefficiency of the current-mode VMM approaches based on RRAMs. A rigorous analysis of different nonideal factors affecting the computational precision indicates that the nonnegligible minimum cell currents, channel length modulation (CLM), and drain-induced barrier lowering (DIBL) are the dominant mechanisms degrading the precision of the proposed VMM. We also show that there exists a tradeoff between the computational precision, dynamic range, and the area-and energy-efficiency of the proposed VMM approach. Therefore, we provide the necessary design guidelines for optimizing the performance. Our preliminary results indicate that an effective computational precision of 6 bits is achievable owing to the inherent compensation effect in the modified 1T-1R blocks. Furthermore, a 4-bit 200 × 200 VMM utilizing the proposed approach exhibits a significantly high energy efficiency of ∼1.5 Pops/J and a throughput of 2.5 Tops/s including the contribution from the input/output (I/O) circuitry. INDEX TERMS 1T-1R array, mixed-signal VMM, time-domain encoding, vector-by-matrix multiplication. I. INTRODUCTION

Long Short-Term Memory Implementation Exploiting Passive RRAM Crossbar Array

“…2. The von-Neumann bottleneck is circumvented in such VMM implementations since the feed-forward propagation (VMM) is performed in-situ at the same location where the neural network parameters (weights) are stored [18]- [20]. Therefore, passive RRAM crossbars are promising candidates for realizing extremely area-and energy-efficient VMM engines.…”

A Bio‐Inspired Neuromorphic Sensory System

“…Such separation of sensing terminals, memories, and central processing units (CPUs) causes large energy consumption, long time latency, and extra hardware costs. [ 4,5 ]…”

“…On the contrary, biological sensing organs, with integrated capabilities of sensing, memory, and computing, acquire and process environmental information in real‐time; therefore, biological sensing organs handle many difficult tasks with low energy consumption and high efficiency. [ 5–9 ] For example, the olfactory system can detect toxic gases in the environment, fire awareness, spoiled food, trigger memories, and also play the biggest role in the sense of flavor. [ 10,11 ] As shown in Figure a, odorants from the environment are captured by the olfactory receptors, which activate the sensory neurons and then generate sparse spike signals.…”

Ultra-Compact Neural Network ADC Exploiting Ferroelectric FET

“…Moreover, ADCs are inevitable while designing compute-in-memory primitives such as inference/multiplyaccumulate (MAC) accelerators based on the cross-point array of emerging non-volatile memories [1]- [5]. ADCs are required at each column of the cross-point array and often dominate the area and energy landscape of such neuromorphic processing engines limiting their area-and energy-efficiency, and the computational precision [6]- [9]. Therefore, there is an urgent need for ultra-compact and energy-efficient ADCs to realize the true potential of the advanced computing and communication systems.…”