• 2019 Update 3
  • 03/07/2019
  • Public Content
  • Download as PDF
Contents

This section describes the Intel® Trace Collector functionality and provides instructions on how to use its features. See the brief overview of each sub-section in the table below.
Section
Description
General instructions on how to trace various types of MPI applications.
Information on tracing MPI events that cause application load imbalance.
Information on tracing non-MPI user-defined events in MPI applications. You can do this automatically for all application functions, or manually only for selected functions or code regions.
Information on how to configure various aspects of Intel® Trace Collector behavior. Configuration is used to enable and disable some Intel Trace Collector functionality and for filtering trace data.
Description of the filtering capabilities of Intel® Trace Collector. Use filtering to trace only information of interest reducing the trace file size and making the results easier to analyze.
Instructions on recording the information about the OpenMP* regions in your application into the trace file.
Information on tracing system input/output calls.
Information on collecting the lightweight statistics about function calls and their communication. Collecting the lightweight statistics is useful for understanding an unknown application.
Instructions on recording the locations of certain functions in the source code. This enables you to easily navigate the source files when analyzing the trace data.
Information on recording the PAPI hardware performance counters.
Information on recording operating system counters, which provide useful information about nodes.
A use case of tracing particular data using various Intel® Trace Collector capabilities. In the example provided, an instrumented application with the use of external libraries is used. Data tracing is presented from the points of view of the application developer and the library developers.
Information on the correctness checking capability. The correctness checker can detect deadlocks, data corruption, and errors with MPI parameters, data types, buffers, communicators, point-to-point messages and collective operations.
Information on tracing distributed applications that work without MPI.

Product and Performance Information

1

Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Notice revision #20110804