《ia-cache-latency-bandwidth-paper》.pdf

  1. 1、本文档共14页,可阅读全部内容。
  2. 2、原创力文档(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
  3. 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载
  4. 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
查看更多
《ia-cache-latency-bandwidth-paper》.pdf

Measuring Cache and Memory Latency and CPU to Memory Bandwidth White Paper Measuring Cache Joshua Ruggiero Computer Systems Engineer Intel Corporation and Memory Latency and CPU to Memory Bandwidth For use with Intel® Architecture December 2008 1 321074 Measuring Cache and Memory Latency and CPU to Memory Bandwidth Executive Summary With today’s computer architecture industry focused on clock speed and processing core count, it is easy to lose sight in other factors that play important roles in overall system performance. Memory latency and bandwidth are two major factors of performance. These two key aspects of system architecture can be broken up into processor cache and system memory. The open source LMBench Benchmarking suite provides utilities to accurately measure latency and bandwidth. Latency and Bandwidth are important measurements to take. Latency and Bandwidth are important measurements to take. These measurements are good indicators that describe the impact during the tuning processes, compare data from one platform to another and possibly identify other potential performance areas of improvement. The open source model provides a solution that adds no extra cost of benchmarking and portable measurements between customers and manufacturer, different teams within the same organization, or even separate members of the same team. §

文档评论(0)

wfkm + 关注
实名认证
内容提供者

该用户很懒,什么也没介绍

1亿VIP精品文档

相关文档