Intel Software Showcase

Introducing the Intel Xeon Phi Coprocesser

The Intel Xeon Phi coprocessor helps to enable dramatic performance gains for the most demanding applications. It takes advantage of familiar and standard programming models to preserve investments while extending hardware support to higher degrees of parallelism with power savings.

Programming the Xeon Phi

Programming the Xeon Phi Intel's tools provide standards-driven parallel programming models that can scale across cores and the techniques used benefit both processors and coprocessors. Programs that utilise multicore processors and many-core coprocessors have a wide variety of options to meet varying needs. These options fully utilise existing widely adopted solutions, such as C, C++, Fortran, OpenMP, MPI and Intel Threading Building Blocks (Intel TBB), and are rapidly driving the development of additional emerging standards such as OpenCL as well as new open entrants such as Intel Cilk Plus.

The graph to the right shows the performance advantage when using a Xeon Phi.

About the Xeon Phi

Breakthrough performance:

Compared with Intel Xeon processor E5 family-based servers, the Intel Xeon Phi coprocessor delivers:

  • Up to 2.5x greater performance on highly parallel applications
  • Up to 3.2x higher peak FLOPS
  • Up to 2.2x higher memory bandwidth
  • Up to 10x better performance on certain financial services applications
  • Up to 4x more performance per watt

Key specifications for 5110P:

  • 60 cores/1.053 GHz/240 threads
  • Up to 1 teraflops double-precision performance
  • 8 Gb memory and 320 Gb/s bandwidth
  • Standard PCIe x16 form factor
  • Linux operating system, IP addressable
  • Supported by the latest Intel software development products
  • 512-bit wide vector engine
  • 32 Kb L1 I/D cache, 512 Kb L2 cache (per core)
  • 8 Gb GDDR5 memory (up to 320 Gb/s)
  • 225W TDP
  • X16 PCIe form factor (requires IA host)
  • Host OS: Red Hat Enterprise Linux 6.x, SuSE Linux 12+
What you need to develop for the Xeon Phi
Intel Parallel Studio XE Cluster Edition

An advanced, comprehensive C ++ and Fortran tool suite that simplifies and speeds high performance computing and cluster application development, debug, and tuning. Scale development efforts with standards-driven compilers, programming models, and tools.

  • Multi-fabric MPI library and advanced MPI error checking and profiling
  • Advanced memory/thread analysis, thread prototyping, and tuning
  • Available with both C++ and Fortran language support or with a single language
Intel System Studio

Intel System Studio provides a complete software tool suite for embedded and mobile system development, comprising debuggers, analysers, compilers and libraries. Supporting Intel Atom, Intel Core and Intel Xeon processors, it helps to accelerate time to market by providing deep hardware and software insights when developing embedded and mobile systems. It enhances code stability with powerful debuggers and highly-optimised maths and media processing libraries and also helps to boost power efficiencies and performance with in-depth system and SoC-wide (System on Chip) power analysis to isolate unexpected sources of power draining. Together these tools accelerate the delivery of next generation, power efficient, reliable embedded and mobile systems.

Learn to develop for the Xeon Phi
Debugging

Introduction to the Debugger - describes the debugger capabilities available for the Xeon Phi.

Command-Line Debugging - guide to debugging Xeon Phi Coprocessor targeted applications on the command-line.

Optimisation

Optimisation - Part 1: Essentials - optimisation and performance tuning guide for Intel Xeon Phi Coprocessors.

Optimisation - Part 2: Hardware Events - optimisation and performance tuning guide for Intel Xeon Phi Coprocessors.

Performance Monitoring Units - guide to using the performance monitoring units available in Xeon Phi.