User Guide

  • 2020
  • 06/18/2020
  • Public Content
Contents

Run Performance Modeling

Run
analyze.py
to predict your application performance on a target device (for example, on a GPU) and generate
Intel® Advisor Beta
output results:
advixe-python <APM>/analyze.py <project-dir> [--options]
where
<APM>
is the
Offload Advisor
environment variable that points to script directory. Replace it with
$APM
on Linux* OS or
%APM%
on Windows* OS.
  • Traditional, not just-in-time (JIT) compiled, Intel® Math Kernel Library (Intel® MKL) code is supported by all analysis types and can be an offload candidate like any other code assuming it is compiled with a general-purpose compiler.
  • Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN) code is JIT compiled and is not supported by the Dependencies analysis. It can be an offload candidate assuming it is compiled with a general-purpose compiler.

Performance Modeling Configuration

By default, Intel® Advisor Beta collections and analysis are optimized to model performance
on an integrated Intel® Processor Graphics Gen11
. You can switch to integrated Intel® Processor Graphics Gen9 configuration by providing an alternate hardware definition model
.toml
file as an input to the
analyze.py
and
collect.py
/
run_oa.py
. For example, for data collection with
collect.py
and performance modeling with
analyze.py
for the Intel® Processor Graphics Gen9 architecture:
advixe-python <APM>/collect.py <project-dir> –-config gen9 [--options] -- <target> [target-options]
advixe-python <APM>/analyze.py <project-dir> [--options]
You can also use the sliders in a
report.html
file to generate a custom hardware configuration file. Refer to Performance Predictor Report.

View Performance Projection Results

Once you have run performance modeling with analyze.py, your results can be found by default in
<advisor-project>/perf_models/mNNNN
. For an overview of the results provided, see Performance Predictor Output Overview.

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