Titlebook: Design and Applications of Emerging Computer Systems; Weiqiang Liu,Jie Han,Fabrizio Lombardi Book 2024 The Editor(s) (if applicable) and T

先鋒派

MOTIF

978-3-031-42480-9The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerl

meditation

Emerging Technologies for Memory-Centric Computingmory-centric paradigm appears to be a good candidate to break the memory and power walls by bringing computation closer to, or even completely inside, the memory. While current approaches are based on conventional volatile memory technologies, further performance improvements can be achieved through

BUCK

An Overview of Computation-in-Memory (CIM) Architectures level of parallelism, and used memory technology. This classification does not only provide an overview of existing CIM architectures with their pros and cons but also unifies the terminology that uniquely identifies these architectures. The chapter also discusses the steps in CIM design flow from

Arresting

Toward Spintronics Non-volatile Computing-in-Memory Architectures using spin devices, including in-memory logic computing and in-memory neural network computing architectures. In this chapter, the focus is mainly on the implementation of in-memory logical/numerical computation using spintronics memory technology and the realization of spintronics neural network

backdrop

Is Neuromorphic Computing the Key to Power-Efficient Neural Networks: A Surveymodern industry theoretically allows unlimited growth; however, in reality, the dreaded power-wall problem in the parallel computing paradigm limits us from exploiting the true potential of AI. Modern neuromorphic accelerators present a lucrative alternative to the traditional artificial neural netw

Dissonance

Electrolysis

An Active Storage System for Intelligent Data Analysis and Managementge, intelligent analysis, and automated management. However, traditional computing-centric storage system design suffers from performance issues caused by data movement and fails to satisfy the demand for intelligent data storage systems. Thereby, we propose an active storage system based on the nea