SPX: Enabling Scalable Synchronizations for General Purpose GPUs

SPX：为通用 GPU 启用可扩展同步

• 批准号: 1725657
• 负责人: Jun Yang
• 金额: $ 85万
• 依托单位: University of Pittsburgh
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-10-01 至 2023-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1725657&HistoricalAwards=false
• 关键词: SPX; Enabling; Scalable; Synchronizations; General

Today?s GPUs have been successfully used for general purpose processing including high-performance computing, machine learning, graph analytics, to name a few applications. However, general purpose computing has different complexity and characteristics from those of graphics processing. A crucial characteristic is the need for coordination and synchronization among parallel threads, which occurs more often in general purpose applications than in graphics rendering. However, unlike common CPU designs, current GPU designs are missing full-fledged support for synchronization primitives that are convenient for implementing transparent scalability and achieving high performance for many parallel applications.This project develops truly effective, efficient, easy-to-use and easy-to-implement synchronization mechanisms for GPGPU while overcoming major obstacles pertaining to the GPU architecture. The first thrust will identify GPGPU synchronization primitives that are deemed sufficient to handle a large set of applications with static and dynamic dependencies. The second thrust will provide solutions and methodologies on how to use the new primitives in the most efficient way to suit the characteristics of applications. The third thrust addresses the main challenges in implementing those primitives in efficiency, possible context switching overhead, and memory capacity limitations. The success in this project will help fuel the spread of accelerator architectures for high-performance computing, cloud and mobile systems. It will increase efficiency for improved performance and reduced energy/power consumption, leading to a greener IT industry. Trained students will become future taskforce to continue the revolution of extending computing into every realm of science, life, commerce and culture.

如今，S GPU已成功用于通用处理，包括高性能计算，机器学习，图形分析，以命名一些应用程序。但是，通用计算与图形处理的复杂性和特征不同。至关重要的特征是平行线程之间需要协调和同步，这比图形渲染更常见。但是，与常见的CPU设计不同，当前的GPU设计缺少对同步基元的全面支持，这些基本可方便地实现透明的可伸缩性和为许多并行应用实现高性能。该项目开发出真正有效，高效，易于使用，易于使用且易于实现的同步机制，用于GPGPU同时访问GPGPU，同时可以依靠GPU访问GPU的范围。第一个推力将识别GPGPU同步基原始人，这些原始基原始人被认为足以处理具有静态和动态依赖性的大量应用程序。第二个推力将提供有关如何以最有效的方式使用新原语的解决方案和方法，以适合应用程序的特征。第三个推力解决了在效率方面实施这些原语的主要挑战，可能的上下文切换开销和内存容量限制。该项目的成功将有助于推动加速器架构的传播，以实现高性能计算，云和移动系统。它将提高效率，以提高性能并减少能源/功耗，从而导致IT行业更加绿色。受过训练的学生将成为未来的工作组，以继续将计算扩展到科学，生活，商业和文化的每个领域的革命。

项目成果

Adaptive Page Migration for Irregular Data-intensive Applications under GPU Memory Oversubscription

GPU内存超额认购下不规则数据密集型应用的自适应页面迁移

• DOI: 10.1109/ipdps47924.2020.00054
• 发表时间: 2020
• 期刊: 2020 IEEE International Parallel and Distributed Processing Symposium (IPDPS
• 作者: Ganguly, Debashis;Zhang, Ziyu;Yang, Jun;Melhem, Rami
• 通讯作者: Melhem, Rami

ModelShield: A Generic and Portable Framework Extension for Defending Bit-Flip based Adversarial Weight Attacks

• DOI: 10.1109/iccd53106.2021.00090
• 发表时间: 2021-10
• 期刊: 2021 IEEE 39th International Conference on Computer Design (ICCD)
• 作者: Yanan Guo;Liang Liu;Yueqiang Cheng;Youtao Zhang;Jun Yang

RFAcc: a 3D ReRAM associative array based random forest accelerator

• DOI: 10.1145/3330345.3330387
• 发表时间: 2019-06
• 期刊: Proceedings of the ACM International Conference on Supercomputing
• 作者: Lei Zhao;Quan Deng;Youtao Zhang;Jun Yang

Aging Capacitor Supported Cache Management Scheme for Solid-State Drives

老化电容支持的固态硬盘缓存管理方案

• DOI: 10.1109/tcad.2019.2949541
• 发表时间: 2020-10
• 期刊: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
• 影响因子: 2.9
• 作者: Gao Congming;Shi Liang;Li Qiao;Liu Kai;Xue Chun Jason;Yang Jun;Zhang Youtao
• 通讯作者: Zhang Youtao

A Dynamic and Proactive GPU Preemption Mechanism Using Checkpointing

使用检查点的动态主动 GPU 抢占机制

• DOI: 10.1109/tcad.2018.2883906
• 发表时间: 2020-01
• 期刊: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
• 影响因子: 2.9
• 作者: Chen Li;Andrew Zigerelli;Jun Yang;Youtao Zhang;Sheng Ma;Yang Guo
• 通讯作者: Yang Guo