Abstract:The bottleneck between the limited memory bandwidth and high speed processing demands is the main cause of problems associated with high volume of data transfers in data-intensive applications. As a possible remedy to these issues, computing-in-memory (CiM) enables a subset of logic and arithmetic operations to be performed where the data resides, i.e., inside the memory. Various CiM designs have been proposed to date, based on different technologies. Given the variety of options available, picking the right design option for a system/application can be a complex task. When choosing a CiM design, it is important to establish evaluation conditions that are as uniform as possible to make a fair choice between available design options. In this paper, we describe a methodology for an uniform benchmarking of CiM designs. Our approach evaluates devices/circuits, arrays and the overall impact of CiM to a system with a framework based on Eva-CiM. As a case study, we analyze the array-level performance of 7 recent CiM designs implemented with SRAM, DRAM, FeFET-RAM, STT-MRAM, SOT-MRAM, and RRAM. After we identify that the FeFET-RAM-based design shows promising energy and delay savings at the array level, we carry out a system level evaluation showing that FeFET-RAM-based CiM outperforms a CMOS SRAM CiM baseline by an average of 60% across a set of 17 benchmarks (with respect to energy savings). Regarding speedups, both technologies offer virtually the same benefit of about 1.5X when compared to a situation where processing does not happen in memory.

Memristor Based Mixed-Precision Computation-in-Memory System.

An Overview of Computing-in-Memory Interfaces

Memristor Based Mixed-Accuracy Computation-in-Memory System.

Modeling and Benchmarking Computing-in-Memory for Design Space Exploration.

Mitigating RC-Delay Induced Accuracy Loss in Analog In-Memory Computing: A Non-Compromising Approach

Mixed-Precision In-Memory Computing

A Memristor Crossbar-Based Computation Scheme with High Precision.

Accuracy Optimization with the Framework of Non-Volatile Computing-In-Memory Systems

Programming Memristor Arrays with Arbitrarily High Precision for Analog Computing.

Emerging Non-Volatile Memories for Computation-in-Memory

Non-volatile Memory Based In-Memory Computing Technology

Memristive Computing Devices and Applications

Design memristor-based computing-in-memory for AI accelerators considering the interplay between devices, circuits, and system

Computing In-Memory, Revisited

Efficient Error-Correcting-Code Mechanism for High-Throughput Memristive Processing-in-Memory

Bulk-Switching Memristor-Based Compute-In-Memory Module for Deep Neural Network Training

Compute-in-Memory for Numerical Computations

Toward Memristive In-Memory Computing: Principles and Applications

C 2 IM: A Compact Computing-In-Memory Unit of 10 Transistors with Standard 6T SRAM

Toward memristive in-memory computing:principles and applications

ICE: Inline Calibration for Memristor Crossbar-Based Computing Engine