arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2012.03112cs.ARcs.DC
收藏

内存中处理的现代导论

A Modern Primer on Processing in Memory

Onur Mutlu, Saugata Ghose, Juan Gómez-Luna, Rachata Ausavarungnirun, Mohammad Sadrosadati, Geraldo F. Oliveira

AI总结:

本文系统梳理了内存中处理(PIM)的研究动机、利用内存处理与近内存处理两类设计方法及其跨层挑战,指出从以处理器为中心向以内存为中心转变是 PIM 实际应用的最大障碍。

摘要:

本文讨论了近期旨在实现靠近数据的计算的研究,我们将这种方法统称为内存中处理(PIM)。PIM 将计算机制放置在数据存储位置内部或附近(即内存芯片或模块内部、3D 堆叠内存的逻辑层中、内存控制器中、存储设备或芯片中),从而减少或消除计算单元与内存/存储单元之间的数据移动。虽然 PIM 的总体思想并不新鲜,但我们讨论了应用以及内存电路和技术中的驱动性趋势,这些趋势极大地加剧了在现代计算系统中启用 PIM 的需求。我们研究了至少两种有前景的 PIM 系统设计新方法,以加速重要的数据密集型应用:(1)利用内存处理(processing-using-memory),它利用内存芯片的基本模拟操作原理在内存中原地执行大规模并行操作;(2)近内存处理(processing-near-memory),它利用不同的逻辑与内存集成技术(例如 3D 堆叠内存技术)将计算逻辑放置在靠近内存电路的位置,从而实现高带宽、低能耗和低延迟的数据访问。在这两种方法中,我们描述并应对了器件、架构、系统、编译器、编程模型和应用中相关的跨层研究、设计与采用挑战。我们的重点在于开发能够以低成本在真实计算平台中采用的 PIM 设计。最后,我们讨论了解决 PIM 实际采用关键挑战的工作。我们相信,从以处理器为中心向以内存为中心的思维模式(和基础设施)转变仍然是 PIM 面临的最大采用挑战,而一旦克服这一挑战,就能释放出一种从根本上高能效、高性能且可持续的计算系统设计、使用和编程新方式。

英文摘要:

This paper discusses recent research that aims to enable computation close to data, an approach we broadly call processing-in-memory (PIM). PIM places computation mechanisms in or near where the data is stored (i.e., inside memory chips or modules, in the logic layer of 3D-stacked memory, in the memory controllers, in storage devices or chips), so that data movement between the computation units and memory/storage units is reduced or eliminated. While the general idea of PIM is not new, we discuss motivating trends in applications as well as memory circuits and technology that greatly exacerbate the need for enabling it in modern computing systems. We examine at least two promising new approaches to designing PIM systems to accelerate important data-intensive applications: (1) processing-using-memory, which exploits fundamental analog operational principles of memory chips to perform massively-parallel operations in-situ in memory, (2) processing-near-memory, which exploits different logic and memory integration technologies (e.g., 3D-stacked memory technology) to place computation logic close to memory circuitry, and thereby enable high-bandwidth, low-energy, and low-latency access to data. In both approaches, we describe and tackle relevant cross-layer research, design, and adoption challenges in devices, architecture, systems, compilers, programming models, and applications. Our focus is on the development of PIM designs that can be adopted in real computing platforms at low cost. We conclude by discussing work on solving key challenges to the practical adoption of PIM. We believe that the shift from a processor-centric to a memory-centric mindset (and infrastructure) remains the largest adoption challenge for PIM, which, once overcome, can unleash a fundamentally energy-efficient, high-performance, and sustainable new way of designing, using, and programming computing systems.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑