AI

The cfsutil ai commands allow you to compile and analyze AI models for Analog Devices embedded targets directly from the command line. Use these commands to build deployable C/C++ code, check model compatibility, profile resource usage, and manage models within a workspace.

cfsutil ai backends — List available AI backends and their supported hardware.
cfsutil ai build — Compile a model into deployable C/C++ source code using a .cfsconfig file or command-line flags.
cfsutil ai clean-cache — Clear the cache of remotely downloaded files.
cfsutil ai compat — Check model compatibility with a target SoC and core.
cfsutil ai model — Add, list, update, and remove AI models within a workspace.
cfsutil ai profile — Profile model resource usage before deployment.
cfsutil ai workspace — Generate a workspace from an AI model file.

Windows dependency

On Windows, cfsutil ai commands require the Visual C++ Redistributable for Visual Studio 2015 to be installed. Install this redistributable manually before running cfsutil ai commands.

Help

Pass --help at any level to view available options:

cfsutil ai --help
cfsutil ai backends --help
cfsutil ai build --help
cfsutil ai clean-cache --help
cfsutil ai compat --help
cfsutil ai model --help
cfsutil ai profile --help
cfsutil ai workspace --help

Global options

The following option is available for all cfsutil ai commands.

Flag	Description
`--json`	Format output as JSON. Affects console output only, not generated files.

Example

cfsutil ai build --soc MAX32690 --core CM4 --model model.tflite --json

Data model search path

The ai build, ai compat, and ai profile commands rely on a .cfsdatamodels index file to look up supported SoCs, cores, and accelerators. A default data model is bundled with the installation, but users can override it by generating their own data model index file. See the CodeFusion Studio GitHub repository for instructions on generating a data model index file.

Use --search-path (-x) to provide an additional directory containing data model files. This flag can be specified multiple times.

Example

cfsutil ai build --search-path /path/to/data-models --soc MAX32690 --core CM4 --model model.tflite