Native and Offload Programming Models for Intel(R) MIC

Using MPI and Xeon Phi™ Offload Together

Overview

Using the Intel® MPI Library along with the offload capabilities of the Intel® Xeon Phi™ coprocessor allows a user to access the capabilities of the coprocessor without the need for direct filesystem access on the coprocessor.  This can lead to administrative benefits by not requiring additional security on the coprocessor's filesystem.

Offloading of MPI functions

Calling MPI functions within an offload region is not supported.

  • 开发人员
  • Linux*
  • C/C++
  • Fortran
  • 高级
  • Intel® MPI Library
  • 消息传递接口
  • Native and Offload Programming Models for Intel(R) MIC
  • 集群计算
  • Intel® Many Integrated Core Architecture
  • Intel® Math Kernel Library (Intel® MKL) on Intel® Many Integrated Core Architecture (Intel® MIC Architecture)

    Starting with Intel® MKL 11.0, support for Intel® Many Integrated Core Architecture (Intel® MIC Architecture) - the Intel® Xeon Phi™ coprocessor has been available. Intel® MKL provides specifically optimized functions for users to take advantages of this new architecture. This page is a collection of online resources on using Intel® MKL on the Intel® Xeon Phi™ coprocessors.

  • 开发人员
  • 教授
  • 学生
  • Linux*
  • 服务器
  • C/C++
  • Fortran
  • 高级
  • 入门级
  • 中级
  • 英特尔® 数学核心函数库
  • MKL MIC
  • MKL Xeon Phi
  • MKL support for Xeon Phi
  • MKL on MIC
  • MKL and Xeon Phi
  • Intel® Xeon Phi
  • MIC
  • Learning Lab
  • Intel MKL Training
  • mkl 11.0
  • MIC architecture
  • Native and Offload Programming Models for Intel(R) MIC
  • 开发工具
  • 大型企业
  • 金融服务行业
  • Intel® Many Integrated Core Architecture
  • 优化
  • 并行计算
  • 订阅 Native and Offload Programming Models for Intel(R) MIC