首页> 外国专利> Mapping calculation units by dividing a calculation model which can be calculated in parallel on an application program

Mapping calculation units by dividing a calculation model which can be calculated in parallel on an application program

机译:通过划分可在应用程序上并行计算的计算模型来映射计算单位

摘要

This invention aims to perform mapping in a user space without concern for an architecture of a parallel computer and to obtain a high- speed mapping pattern. An N-dimensional model is divided by a user into a plurality of calculation units. The calculation units and physical processors are assigned with an identification code. Correspondence therebetween is held on an address conversion table. By accessing the address conversion table, the user can freely perform a mapping operation. Thereafter, interprocessor communication is carried out by the use of the identification code assigned to the calculation units.
机译:本发明旨在在用户空间中执行映射而无需考虑并行计算机的体系结构,并获得高速映射模式。用户将N维模型划分为多个计算单元。计算单元和物理处理器分配有识别码。它们之间的对应关系保存在地址转换表上。通过访问地址转换表,用户可以自由地执行映射操作。此后,通过使用分配给计算单元的识别码进行处理器间通信。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

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