细粒度同步的Warp调度

2018 IEEE International Symposium on High Performance Computer Architecture (HPCA) Pub Date : 2018-02-01 DOI:10.1109/HPCA.2018.00040

Ahmed Eltantawy, Tor M. Aamodt

{"title":"细粒度同步的Warp调度","authors":"Ahmed Eltantawy, Tor M. Aamodt","doi":"10.1109/HPCA.2018.00040","DOIUrl":null,"url":null,"abstract":"Fine-grained synchronization is employed in many parallel algorithms and is often implemented using busy-wait synchronization (e.g., spin locks). However, busy-wait synchronization incurs significant overheads and existing CPU solutions do not readily translate to single-instruction, multiple-thread (SIMT) graphics processor unit (GPU) architectures. In this paper, we propose Back-Off Warp Spinning (BOWS), a hardware warp scheduling policy that extends existing warp scheduling policies to temporarily deprioritize warps executing busy wait code. In addition, we propose Dynamic Detection of Spinning (DDOS), a novel hardware mechanism for accurately and efficiently detecting busy-wait synchronization on GPUs. On a set of GPU kernels employing busy-wait synchronization, DDOS identifies all busy-wait loops incurring no false detections. BOWS improves performance by 1.5× and reduces energy consumption by 1.6× versus Criticality-Aware Warp Acceleration (CAWA) [14].,,,,","PeriodicalId":154694,"journal":{"name":"2018 IEEE International Symposium on High Performance Computer Architecture (HPCA)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"23","resultStr":"{\"title\":\"Warp Scheduling for Fine-Grained Synchronization\",\"authors\":\"Ahmed Eltantawy, Tor M. Aamodt\",\"doi\":\"10.1109/HPCA.2018.00040\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Fine-grained synchronization is employed in many parallel algorithms and is often implemented using busy-wait synchronization (e.g., spin locks). However, busy-wait synchronization incurs significant overheads and existing CPU solutions do not readily translate to single-instruction, multiple-thread (SIMT) graphics processor unit (GPU) architectures. In this paper, we propose Back-Off Warp Spinning (BOWS), a hardware warp scheduling policy that extends existing warp scheduling policies to temporarily deprioritize warps executing busy wait code. In addition, we propose Dynamic Detection of Spinning (DDOS), a novel hardware mechanism for accurately and efficiently detecting busy-wait synchronization on GPUs. On a set of GPU kernels employing busy-wait synchronization, DDOS identifies all busy-wait loops incurring no false detections. BOWS improves performance by 1.5× and reduces energy consumption by 1.6× versus Criticality-Aware Warp Acceleration (CAWA) [14].,,,,\",\"PeriodicalId\":154694,\"journal\":{\"name\":\"2018 IEEE International Symposium on High Performance Computer Architecture (HPCA)\",\"volume\":\"45 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-02-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"23\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2018 IEEE International Symposium on High Performance Computer Architecture (HPCA)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/HPCA.2018.00040\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 IEEE International Symposium on High Performance Computer Architecture (HPCA)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/HPCA.2018.00040","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 23

摘要

细粒度同步在许多并行算法中使用，并且通常使用忙等待同步(例如，自旋锁)来实现。然而，忙碌等待同步带来了巨大的开销，并且现有的CPU解决方案不容易转换为单指令、多线程(SIMT)图形处理器单元(GPU)架构。在本文中，我们提出了一种硬件Warp调度策略，它扩展了现有的Warp调度策略，以暂时降低执行繁忙等待代码的Warp的优先级。此外，我们提出了动态检测旋转(DDOS)，这是一种新的硬件机制，可以准确有效地检测gpu上的忙等待同步。在一组采用忙等待同步的GPU内核上，DDOS识别所有不会产生错误检测的忙等待循环。与临界感知曲速加速(CAWA)相比，bow提高了1.5倍的性能，降低了1.6倍的能耗[14].，，，，

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Warp Scheduling for Fine-Grained Synchronization

Fine-grained synchronization is employed in many parallel algorithms and is often implemented using busy-wait synchronization (e.g., spin locks). However, busy-wait synchronization incurs significant overheads and existing CPU solutions do not readily translate to single-instruction, multiple-thread (SIMT) graphics processor unit (GPU) architectures. In this paper, we propose Back-Off Warp Spinning (BOWS), a hardware warp scheduling policy that extends existing warp scheduling policies to temporarily deprioritize warps executing busy wait code. In addition, we propose Dynamic Detection of Spinning (DDOS), a novel hardware mechanism for accurately and efficiently detecting busy-wait synchronization on GPUs. On a set of GPU kernels employing busy-wait synchronization, DDOS identifies all busy-wait loops incurring no false detections. BOWS improves performance by 1.5× and reduces energy consumption by 1.6× versus Criticality-Aware Warp Acceleration (CAWA) [14].,,,,

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

2018 IEEE International Symposium on High Performance Computer Architecture (HPCA)

自引率

0.00%

发文量