Multi-Layer In-Memory Processing
Multi-Layer In-Memory Processing
复制标题
DOI：
10.1109/micro56248.2022.00068
复制
发表时间：
2022-10
期刊：
影响因子：
--
通讯作者：
Daichi Fujiki;Alireza Khadem;S. Mahlke;R. Das
中科院分区：
文献类型：
--
作者：
Daichi Fujiki;Alireza Khadem;S. Mahlke;R. Das
内存计算通过融合内存和计算，为计算机架构带来了革命性的变革，使数据密集型计算能够减少数据通信。尽管内存计算在内存层次结构的每一层都取得了有前景的成果，但对于具有多个可计算内存的系统的综合方法尚未得到研究。本文提出了一种整体的、由应用驱动的构建多层内存内处理（MLIMP）系统的方法，使具有不同计算需求的应用能够在集成的MLIMP系统中受益于异构计算资源。通过将并发任务调度引入MLIMP，我们提高了图神经网络以及数据并行应用多程序设计的性能和能效。
In-memory computing provides revolutionary changes to computer architecture by fusing memory and computation, allowing data-intensive computations to reduce data communications. Despite promising results of in-memory computing in each layer of the memory hierarchy, an integrated approach to a system with multiple computable memories has not been examined. This paper presents a holistic and application-driven approach to building Multi-Layer In-Memory Processing (MLIMP) systems, enabling applications with variable computation demands to reap the benefits of heterogeneous compute resources in an integrated MLIMP system. By introducing concurrent task scheduling to MLIMP, we achieve improved performance and energy efficiency for graph neural networks and multiprogramming of data parallel applications.