PIUMA: Programmable Integrated Unified Memory Architecture
PIUMA: Programmable Integrated Unified Memory Architecture
复制标题
DOI：
--
复制
发表时间：
2020-10
期刊：
影响因子：
--
通讯作者：
S. Aananthakrishnan;Nesreen Ahmed;Vincent Cavé;Marcelo H. Cintra;Y. Demir;Kristof Du Bois;Stijn Eyerman;J. Fryman;Ivan B. Ganev;W. Heirman;Hans-Christian Hoppe;J. Howard;I. Hur;Midhunchandra Kodiyath;Samkit Jain;D. Klowden;M. Landowski;L. Montigny;A. More;Przemysław Ossowski;R. Pawlowski;Nick Pepperling;F. Petrini;Mariusz Sikora;B. Seshasayee;Shaden Smith;Sebastian Szkoda;Sanjaya Tayal;Jesmin Jahan Tithi;Yves Vandriessche;Izajasz P. Wrosz
中科院分区：
文献类型：
--
作者：
S. Aananthakrishnan;Nesreen Ahmed;Vincent Cavé;Marcelo H. Cintra;Y. Demir;Kristof Du Bois;Stijn Eyerman;J. Fryman;Ivan B. Ganev;W. Heirman;Hans-Christian Hoppe;J. Howard;I. Hur;Midhunchandra Kodiyath;Samkit Jain;D. Klowden;M. Landowski;L. Montigny;A. More;Przemysław Ossowski;R. Pawlowski;Nick Pepperling;F. Petrini;Mariusz Sikora;B. Seshasayee;Shaden Smith;Sebastian Szkoda;Sanjaya Tayal;Jesmin Jahan Tithi;Yves Vandriessche;Izajasz P. Wrosz
高性能的大规模图分析对于及时分析大数据集中的关系至关重要。传统的处理器架构遭受低效的资源使用和对图工作负载的不良缩放。为了实现高效和可扩展的图形分析，英特尔开发了可编程集成统一内存架构（PIUMA）。PIUMA由许多多线程内核、细粒度内存和网络访问、全局共享地址空间和强大的卸载引擎组成。本文介绍了PIUMA体系结构，并提供了初步的性能估计，预计PIUMA节点将优于传统的计算节点的一到两个数量级。此外，PIUMA继续跨多个节点扩展，这在传统的多节点设置中是一个挑战。
High performance large scale graph analytics is essential to timely analyze relationships in big data sets. Conventional processor architectures suffer from inefficient resource usage and bad scaling on graph workloads. To enable efficient and scalable graph analysis, Intel developed the Programmable Integrated Unified Memory Architecture (PIUMA). PIUMA consists of many multi-threaded cores, fine-grained memory and network accesses, a globally shared address space and powerful offload engines. This paper presents the PIUMA architecture, and provides initial performance estimations, projecting that a PIUMA node will outperform a conventional compute node by one to two orders of magnitude. Furthermore, PIUMA continues to scale across multiple nodes, which is a challenge in conventional multinode setups.