FoMR: Collaborative Research: Dependent ILP: Dynamic Hoisting and Eager Scheduling of Dependent Instructions

FoMR：协作研究：相关 ILP：相关指令的动态提升和紧急调度

• 批准号: 1823417
• 负责人: David Whalley
• 金额: $ 21.5万
• 依托单位: Florida State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-10-01 至 2022-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1823417&HistoricalAwards=false
• 关键词: FoMR; Collaborative; Research; Dependent; ILP

Instruction-level parallelism (ILP) in computing allows different machine-level instructions within an application to execute in parallel within a micro-processor. Exploitation of ILP has provided significant performance benefits in computing, but there has been little improvement in ILP in recent years. This project proposes a new approach called "eager execution" that could significantly increase ILP. The success of many applications depends on how efficiently they can be executed. The proposed eager execution technique will benefit applications that span those running on mobile devices to large data applications running on the ever-growing number of data centers. Enabling better systems at all scales will further enable the ubiquitous computing that continues to pervade lives.The project's approach includes the following advantages: (1) immediately-dependent consumer instructions can be more quickly delivered to functional units for execution; (2) the execution of instructions whose source register values have not changed since its last execution can be detected and redundant computation can be avoided; (3) the dependency between a producer/consumer pair of instructions can sometimes be collapsed so they can be simultaneously dispatched for execution; (4) consumer instructions from multiple paths may be speculatively executed and their results can be naturally retained in the paradigm to avoid re-execution after a branch misprediction; and (5) critical instructions can be eagerly executed to improve performance, which include loads to prefetch cache lines and pre-computation of branch results to avoid branch misprediction delays.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

计算中的指令级并行（ILP）允许应用程序中的不同机器级指令在微处理器中并行执行。ILP的开发在计算中提供了显著的性能优势，但近年来ILP的改进很少。该项目提出了一种名为“渴望执行”的新方法，可以显着提高ILP。许多应用程序的成功取决于它们的执行效率。所提出的急切执行技术将使应用程序受益，所述应用程序跨越在移动的设备上运行的应用程序到在不断增长的数据中心上运行的大型数据应用程序。该项目的方法包括以下优点：（1）直接依赖的消费者指令可以更快地交付给功能单元执行;（2）可以检测到自上次执行以来源寄存器值没有改变的指令的执行，并可以避免冗余计算;（3）可以检测到自上次执行以来源寄存器值没有改变的指令的执行。（3）生产者/消费者指令对之间的依赖性有时可以被折叠，因此它们可以被同时分派以供执行;（4）来自多个路径的消费者指令可以被推测性地执行，并且它们的结果可以自然地保留在范例中以避免在分支误预测之后的重新执行;以及（5）可以急切地执行关键指令以提高性能，其包括预取高速缓存行的加载和预取高速缓存行的加载。计算分支结果，以避免分支误预测延迟。该奖项反映了NSF的法定使命，并已被认为是值得通过评估使用基金会的智力价值和更广泛的影响审查标准。

项目成果

Facilitating the Bootstrapping of a New ISA

• DOI: 10.1145/3589610.3596282
• 发表时间: 2023-06
• 期刊: Proceedings of the 24th ACM SIGPLAN/SIGBED International Conference on Languages, Compilers, and Tools for Embedded Systems
• 作者: Abigail Mortensen;Scott Pomerville;D. Whalley;Soner Önder;Gang-Ryung Uh

The Domestic Computer Science Graduate Students Are There, We Just Need to Recruit Them

国内计算机研究生已经有了，我们只需要招就可以了

• DOI: 10.1145/3448249
• 发表时间: 2021
• 期刊: Communications of the ACM
• 影响因子: 22.7
• 作者: Whalley, David;Yuan, Xin;Liu, Xiuwen
• 通讯作者: Liu, Xiuwen

Decreasing the Miss Rate and Eliminating the Performance Penalty of a Data Filter Cache

降低丢失率并消除数据过滤器缓存的性能损失

• 发表时间: 2021
• 期刊: ACM transactions on architecture and code optimization
• 影响因子: 1.6
• 作者: Stokes, Michael;Whalley, David;Onder, Soner
• 通讯作者: Onder, Soner

Improving Energy Efficiency by Memoizing Data Access Information

• DOI: 10.1109/islped.2019.8824951
• 发表时间: 2019-07
• 期刊: 2019 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED)
• 作者: Michael Stokes;Ryan Baird;Zhaoxiang Jin;D. Whalley;Soner Önder