Fastpath Speculative Parallelization

机译：快速路径投机并行化

获取原文

获取原文并翻译 | 示例

页面导航

摘要
著录项
相似文献
相关主题

摘要

We describe Fastpath, a system for speculative parallelization of sequential programs on conventional multicore processors. Our system distinguishes between the lead thread, which executes at almost-native speed, and speculative threads, which execute somewhat slower. This allows us to achieve nontrivial speedup, even on two-core machines. We present a mathematical model of potential speedup, parameterized by application characteristics and implementation constants. We also present preliminary results gleaned from two different Fastpath implementations, each derived from an implementation of software transac-tional memory.

机译：我们描述了Fastpath，这是一种用于在常规多核处理器上对顺序程序进行推测并行化的系统。我们的系统区分执行速度几乎为本地的引导线程和执行速度较慢的推测线程。这使我们即使在两核计算机上也可以实现非凡的加速。我们提出了潜在加速的数学模型，并由应用程序特征和实现常数进行了参数化。我们还介绍了从两种不同的Fastpath实现中收集的初步结果，每种实现均来自软件事务存储的实现。

著录项

来源
《Languages and compilers for parallel computing》|2009年|p.338-352|共15页
会议地点 Newark DE(US);Newark DE(US)
作者
Michael F. Spear; Kirk Kelsey; Tongxin Bai; Luke Dalessandro; Michael L. Scott; Chen Ding; Peng Wu;
展开▼
作者单位

Lehigh University;

University of Rochester;

University of Rochester;

University of Rochester;

University of Rochester;

University of Rochester;

IBM T. J. Watson Research Center;

展开▼
会议组织
原文格式 PDF
正文语种 eng
中图分类计算技术、计算机技术;
关键词

相似文献

外文文献
中文文献
专利

1. The LRPD test: speculative run-time parallelization of loops with privatization and reduction parallelization [J] . Rauchwerger L., Padua D.A. IEEE Transactions on Parallel and Distributed Systems . 1999,第2期

机译：LRPD测试：具有私有化和约简并行化的循环的推测性运行时并行化
2. The LRPD test: speculative run-time parallelization of loops with privatization and reduction parallelization [J] . Lawrence Rauchwerger, David Padua ACM SIGPLAN Notices: A Monthly Publication of the Special Interest Group on Programming Languages . 1995,第6期

机译：LRPD测试：具有私有化和约简并行化的循环的推测性运行时并行化
3. Using the Xeon Phi Platform to Run Speculatively-Parallelized Codes [J] . Alvaro Estebanez, Diego R. Llanos, Arturo Gonzalez-Escribano International journal of parallel programming . 2017,第2期

机译：使用Xeon Phi平台运行推测性并行代码
4. Harmonizing Speculative and Non-Speculative Execution in Architectures for Ordered Parallelism [C] . Mark C. Jeffrey, Victor A. Ying, Suvinay Subramanian, Annual IEEE/ACM International Symposium on Microarchitecture . 2018

机译：协调架构中的有序和无序执行以实现有序并行
5. A Sensible Approach to Speculative Automatic Parallelization [D] . Apostolakis, Sotiris. 2021

机译：一种可笑的自动并行化方法
6. Review of the speculative role of co-infections in [O] . Milan R. Obradovic, Mariela Segura, Joaquim Segalés, 2021

机译：审查共感染的投机作用
7. Fastpath speculative parallelization [O] . Michael F. Spear, Kirk Kelsey, Tongxin Bai, 2010

机译：快速路径推测并行化

Fastpath Speculative Parallelization

摘要

著录项

相似文献

相关主题

期刊订阅