doc/develop/manifest/external/zephelin.rst
.. _external_module_zephelin:
Zephelin ########
Introduction
Zephyr Profiling Library_ (ZPL), or Zephelin for short, is a library which enables capturing
and reporting runtime performance metrics for profiling and detailed analysis of Zephyr
applications, with a special focus on applications running AI/ML inference workloads.
In addition to the above, Zephelin also simplifies the analysis of AI runtimes, such as LiteRT_
and microTVM_, allowing you to gain a better understanding of the underlying bottlenecks or
potential opportunities for optimization.
Zephelin features:
Tracing execution of Zephyr applications on hardware
Obtaining traces using such backends as UART, USB or debug adapter
Delivering traces in CTF and TEF formats
Scripts for capturing traces from device
Collecting readings from:
Displaying details on executed neural network layers in the LiteRT or microTVM runtime:
Compilation-level and runtime-level configuration of the library
Ability to configure a profiling tier, controlling the subsystems and the amount of data collected
All of the above can be analyzed with Zephelin Trace Viewer_.
Usage With Zephyr
To use Zephelin as a Zephyr :ref:module <modules>, add the following entry:
.. code-block:: yaml
manifest: projects: - name: zephelin url: https://github.com/antmicro/zephelin revision: main path: modules/zephelin # adjust the path as needed
to a Zephyr submanifest (e.g. zephyr/submanifests/zephelin.yaml) and run west update, or
add it as a West project in your project's west.yaml manifest.
Please consult the Zephelin documentation_ for more information.
References
.. target-notes::
.. _Zephyr Profiling Library: https://github.com/antmicro/zephelin
.. _Zephelin documentation: https://antmicro.github.io/zephelin/
.. _Zephelin Trace Viewer: https://antmicro.github.io/zephelin-trace-viewer
.. _LiteRT: https://ai.google.dev/edge/litert
.. _microTVM: https://tvm.apache.org/docs/v0.9.0/topic/microtvm/index.html