Reconfigurable Compute-In-Memory on Field-Programmable Ferroelectric Diodes

Abstract

The deluge of sensors and data generating devices has driven a paradigm shift in modern computing from arithmetic-logic centric to data centric processing. At a hardware level, this presents an urgent need to integrate dense, high-performance and low-power memory units with Si logic-processor units. However, data-heavy problems such as search and pattern matching also require paradigm changing innovations at the circuit and architecture level to enable compute in memory (CIM) operations. CIM architectures that combine data storage yet concurrently offer low-delay and small footprint are highly sought after but have not been realized. Here, we present Aluminum Scandium Nitride (AlScN) ferroelectric diode (FeD) memristor devices that allow for storage, search and neural network-based pattern recognition in a transistor-free architecture. Our devices can be directly integrated on top of Si processors in a scalable, back-end-of-line process. We leverage the field-programmability, non-volatility and non-linearity of FeDs to demonstrated circuit blocks that can support search operations in-situ memory with search delay times < 0.1 ns and a cell footprint < 0.12 µm2. In addition, we demonstrate matrix multiplication operations with 4-bit operation of the FeDs. Our results highlight FeDs as promising candidates for fast, efficient, and multifunctional CIM platforms.

Sub‐Microsecond Polarization Switching in (Al,Sc)N Ferroelectric Capacitors Grown on Complementary Metal–Oxide–Semiconductor‐Compatible Aluminum Electrodes

“…[ 10 ] These attributes are expected to benefit the on/off current ratio and memory window for NVM devices. [ 11 ] Ferroelectric switching was recently reported in sub‐100 nm thick films deposited below 400 °C on platinized silicon wafers [10b,c,12] . These studies suggest that (Al,Sc)N is a promising material for ferroelectric memories, and the low deposition temperatures make these materials compatible with the integration of (Al,Sc)N‐embedded NVMs directly into modern CMOS‐BEOL processes.…”

Ultrathin Nitride Ferroic Memory with Large ON/OFF Ratios for Analog In‐Memory Computing

“…[40][41][42][43][44][45][46][47][48] In this regard, a few memristor demonstrations based on ScAlN have been reported, yet the ScAlN layer employed are thick (>20 nm) which causes high operation voltages and potentially scalability problems. [49][50][51][52] Moreover, the ferroelectric synaptic weight update as well as precise VMM operation and image processing in ultrathin ScAlN has remained elusive.…”

Computational understanding role of vacancies and distortions in wurtzite ferroelectric memory materials: implications for device miniaturization

“…50 The large leakage current causes problems in the ferroelectric extraction/measurement methodologies, such as the two-frequency technique and pulse-up/pulse-down strategy, which are harnessed to minimize the leakage currents. 51,52 Moreover, the trade-off between decreasing leakage current, and simultaneously, minimizing thin-film thickness represents an impediment to fulfilling the promise of FEM. 53 Hence, clarification of possible origin of the high leakage current is needed.…”