Intel® Thread Checker

Are High School Whiz Kids Ready to "Think Parallel?"

Can technical high school students solve complex parallel programming challenges, typically reserved for advanced university courses? A three-day boot-camp hosted by the Intel® Academic Community and Brooklyn Technical High School with industry experts.
  • Professors
  • Students
  • Game Development
  • Intel® Thread Checker
  • Parallel Computing
  • Intel® Thread Checker limited support with statically linked C++ runtime library

    Intel Thread Checker doesn't support applications which have statically linked C++ runtime libraries. In order to support C++ exception propagation in the instrumented binary image, Thread Checker requires __cxa_rethrow and __cxa_begin_catch functions to be exported. These functions are not exported in statically linked applications.

    If the C++ runtime library is statically linked in an application, Thread Checker will stop the analysis with the following error message:

  • Linux*
  • Intel® Thread Checker
  • runtime
  • Performance Tools for Software Developers - "License Expired" message received even though you have a valid license

    After installing Intel® Thread Profiler for Windows* or Intel® Thread Checker for Windows*, when you try to run an Intel® VTune™ Performance Analyzer Wizard, you get a "License Expired" message; even though you have a valid license (e.g. just obtained an Evaluation license).

    To resolve this, click the drop-down box and choose Threading Wizards, and select an available wizard. Analyzer wizards refer to Intel VTune Performance Analyzer wizards, which are not covered by the Intel Thread Profiler or Intel Thread Checker license.

    OR

  • Microsoft Windows* (XP, Vista, 7)
  • Intel® Thread Checker
  • OpenMP, Linux и немного фана

    Попался вот такой кусок кода для вычисления тривиального числа Pi

    n = 1000000000
    62    h   = 1.0 / (double) n;
    63    sum = 0.0;
    64
    65    for (i = 1; i <= n; i++)
    66    {
    67           x = h * ((double)i - 0.5);
    68           sum +=(4.0 / (1.0 + x*x));
    69    }
    70
    71    mypi = h * sum;

    изначально все строилось для MPI , правда захотелось перенести на OpenMP

    казалось бы все просто:

     53 #pragma omp parallel for reduction(+: sum)
     54 for (int i = 1; i <= n; i++)

    Subscribe to Intel® Thread Checker