首页> 外文期刊>International Journal of High Performance Computing Applications >IMPROVING PERFORMANCE BY REDUCING THE MEMORY FOOTPRINT OF SCIENTIFIC APPLICATIONS
【24h】

IMPROVING PERFORMANCE BY REDUCING THE MEMORY FOOTPRINT OF SCIENTIFIC APPLICATIONS

机译:通过减少科学应用程序的内存足迹来提高性能

获取原文
获取原文并翻译 | 示例
           

摘要

Over the last two decades, processor speeds have improved much faster than memory speeds. As a result, memory access delay is a major performance bottleneck in today's systems. Compilers often fail to choreograph data and computation automatically to avoid memory access delay; we have developed an annotation-driven source-to-source transformation tool for this purpose. This tool uses a set of compiler transformations that improve temporal reuse in scientific applications (1) by reducing the size of temporary arrays and (2) by overlaying storage for multiple temporary arrays that are not live at the same time. We also describe two supporting transformations, statement motion and loop alignment, that improve the effectiveness of storage reduction. Our experiments with a numerical kernel and two weather codes show that our storage reduction optimizations amplify the benefits of loop transformations and double performance achievable with loop transformations alone.
机译:在过去的二十年中,处理器速度的提高速度远快于内存速度。因此,内存访问延迟是当今系统中的主要性能瓶颈。编译器通常无法自动编排数据和计算以避免内存访问延迟。为此,我们已经开发了注释驱动的源到源转换工具。该工具使用了一组编译器转换,可改善科学应用程序中的时间重用性(1)通过减小临时数组的大小,以及(2)通过覆盖不同时存在的多个临时数组的存储来实现。我们还描述了两种支持的转换,即语句运动和循环对齐,可提高存储减少的效率。我们使用数字内核和两个天气代码进行的实验表明,我们的存储减少优化功能扩大了循环转换的好处,并且仅循环转换就可以实现双重性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号