Machine Learning (Silicon Labs AI/ML) Extension SDK Version 2.1.0 (June 18, 2025) - Release Notes#

Simplicity SDK Version 2025.6.0

The Machine Learning Extension is also known as "Silicon Labs AI/ML" and is provided as an extension to the Simplicity SDK. It enables AI/ML development on Series 2 and SiWG917 devices using the Tensorflow Lite for Microcontrollers (TFLM) framework.

Click here for earlier releases.

Release Summary#

Key Features | API Changes | Bug Fixes | Chip Enablement

Key Features#

Source code moved to an Extension.
AI/ML Extension is now Generally Available (GA), it was released as an alpha build earlier.
Added support for SiWG917 chip family through software optimizations.
Machine Learning (AI/ML) developer documentation has been re-organized.

API Changes#

New APIs for model invocation and execution for SiWG917 chip family.
New variables for interacting with the ML models.

Bug Fixes#

Fixed: memory leak of mic when model does not load correctly.

Chip Enablement#

EFR32xG2x
SiWG917

Key Features#

New Features | Enhancements | Removed Features | Deprecated Features

Note: See Feature Matrix for a list of any appliable APIs, examples, software variants, modes, hardware, and host interfaces applicable for each feature.

New Features#

Source code moved into AI/ML Extension.
Audio applications enabled for SiWG917 chip family, see Audio Classifier and Voice Control Light example applications for SiWG917.
New APIs added for model invocation and execution for SiWG917 chip family, see API Changes for details.
Machine Learning (AI/ML) developer documentation has been re-organized and revamped completely to harmonize it with documentation best practices, check for new paths in the documentation.

Enhancements#

Only new features have been added in this release. No enhancements were made to existing features.

Removed Features#

None.

Deprecated Features#

None.

API Changes#

New APIs | Modified APIs | Removed APIs | Deprecated APIs

New APIs#


New API Signature	Deprecated API replaced by this (if any)
sl_status_t sl_ml_<model_name>_model_init()	None
sl_status_t sl_ml_<model_name>_model_run()	None
static TfliteMicroModel <model_name>_model	None
static sl_status_t <model_name>_model_status	None
uint8_t* <model_name>_model_flatbuffer	None
const int <model_name>_model_flatbuffer_length	None

Modified APIs#

None.

Removed APIs#

None.

Deprecated APIs#

None.

Bug Fixes#

ID	Issue Description	GitHub / Salesforce Reference (if any)	Affected Software Variants, Hardware, Modes, Host Interfaces
1452807	Memory Leak in `sl_ml_audio_feature_generation_init()` due to Missing Resource Cleanup	None	Standard EFR32xG2x SoC

Chip Enablement#

Chip Family	OPNs / Boards / OPN Combinations	Supported Software Variants (if applicable)	Supported Modes	Supported Host Interfaces
Chip	OPN: EFR32xG2x, SiWG917x Boards: BRD2601b, BRD2608a, BRD2605a, BRD2705a External Hosts: N/A	Standard	SoC

Application Example Changes#

New Examples | Modified Examples | Removed Examples | Deprecated Examples

New Examples#

Example Name	Description	Supported Software Variants (if applicable)	Supported Modes	Supported OPNs / Boards / OPN Combinations	Supported Host Interfaces
Audio Classifier for SiWG917 See readme.md	This application uses TensorFlow Lite for Microcontrollers to run audio classification machine learning models to classify words from audio data recorded from a microphone. The detection is visualized using the LED's on the board and the classification results are written to the VCOM serialport.	Standard	SoC	OPN: SiWG917x Boards: BRD2605a External Hosts: N/A
Voice Control Light for SiWG917 See readme.md	This application uses TensorFlow Lite for Microcontrollers to detect the spoken words "on" and "off" from audio data recorded on the microphone. The detected keywords are used to control an LED on the board.	Standard	SoC	OPN: SiWG917x Boards: BRD2605a External Hosts: N/A

Example Name

Description

Supported Software Variants (if applicable)

Supported Modes

Supported OPNs / Boards / OPN Combinations

Supported Host Interfaces

Audio Classifier for SiWG917

See readme.md

This application uses TensorFlow Lite for Microcontrollers to run audio classification machine learning models to classify words from audio data recorded from a microphone. The detection is visualized using the LED's on the board and the classification results are written to the VCOM serialport.

Standard

OPN: SiWG917x
Boards: BRD2605a
External Hosts: N/A

Voice Control Light for SiWG917

See readme.md

This application uses TensorFlow Lite for Microcontrollers to detect the spoken words "on" and "off" from audio data recorded on the microphone. The detected keywords are used to control an LED on the board.

Standard

OPN: SiWG917x
Boards: BRD2605a
External Hosts: N/A

Modified Examples#

None.

Removed Examples#

None.

Deprecated Examples#

None.

Known Issues and Limitations#

ID	Issue or Limitation Description	GitHub / Salesforce Reference (if any)	Workaround (if any)	Affected Software Variants, Hardware, Modes, Host Interfaces
1463269	Blink app behaves differently when running on Series 2, compared to SiWG917. The LED has a baseline brightness on SiWG917 while blinking but it turns off completely on Series 2. This is due to the different LED driver implementations on the two chip families.	None	A workaround has not been implemented yet. The issue is under investigation to transition from a software PWM to a hardware PWM implementation for the LED driver on SiWG917.	SiWG917M111MGTBA BRD2605a SoC
1463268	Audio feature generation component shows config error for SiWG917.	None	This issue is cosmetic and a fix is under investigation for a future release. This issue does not affect Series 2 projects.	SiWG917M111MGTBA BRD2605a SoC
1464105	API calls for model invocation and execution are different on Series 2 and SiWG917 chip families.	None	The Series 2 uses `sl_ml_model_init()`, and `sl_tflite_micro_get_interpreter()->Invoke()` for model invocation and execution, respectively, whereas SiWG917 uses `sl_ml_<model_name>_model_init()`, `sl_ml_<model_name>_model_run()`.	SiWG917M111MGTBA BRD2605a SoC

Impact of Release Changes#

Impact Statements | Migration Guide

Impact Statements#

Change	Impact	Affected Software Variants if applicable	Affected Modes	Affected OPNs / Boards / OPN Combinations	Affected Host Interfaces
Source code moved into AI/ML Extension	Components moved from "Machine Learning" to "Silicon Labs AI/ML v2.1.0 > Machine Learning"	Standard	SoC	OPN: EFR32xG2x, SiWG917x Boards: BRD2601b, BRD2608a, BRD2605a, BRD2705a External Hosts: N/A

Migration Guide#

Migrate to AI/ML Extension

Using This Release#

What's in the Release? | Compatible Software | Installation and Use | Help and Feedback

What's in the Release?#

This is the first version where AI/ML software is provided as an extension to the Simplicity SDK. It supports the Series 2 and SiWG917 family of Silicon Labs devices using the Tensorflow Lite for Microcontrollers (TFLM) framework. The highlight of this release are the performance improvements to the SDK, bringing support for SiWG917 chip family, which has a vastly different SoC architecture compared to the Series 2 devices. SiWG917 is a high-performance, low-power Wi-Fi chip that is ideal for AI/ML applications. Hence, your AI/ML applications can now leverage Wi-Fi connectivity for cloud-based AI/ML applications and services.

Compatible Software#

Software	Compatible Version or Variant
Software Development Kit (SDK)	Simplicity SDK: 2025.6.0 WiSeConnect SDK: 3.5.0

Installation and Use#

To upgrade your existing software with this release, update Simplicity Studio to the latest, SiSDK to v2025.6.0, WiSeConnect SDK to v3.5.0, and AI/ML Extension to v2.1.0 from Studio installation manager, or download the SDKs from the respective links listed in Compatible Software section above. To update AI/ML Extension, please refer to AI/ML Extension Setup guide.

To run your first demo, see our Getting Started

To kick start your development, see our Developer's Guide

For information about Secure Vault Integration, see Secure Vault.

To review Security and Software Advisory notifications and manage your notification preferences:

Go to https://community.silabs.com/.
Log in with your account credentials.
Click your profile icon in the upper-right corner of the page.
Select Notifications from the dropdown menu.
In the Notifications section, go to the My Product Notifications tab to review historical Security and Software Advisory notifications
To manage your preferences, use the Manage Notifications tab to customize which product updates and advisories you receive.

To learn more about the software in this release, dive into our online documentation

Help and Feedback#

Contact Silicon Labs Support.
To use our Ask AI tool to get answers, see the search field at the top of this page.
Note: Ask AI is experimental.
Get help from our developer community.

Feature Matrix#

Supported Features | Unsupported Features

Supported Features#

Feature Name	Description	Quality	Related API Names	Supported Software Variants, Hardware, Modes, Host Interfaces	Related Example Names
Model specific APIs	Model specific APIs for invoking and executing AI/ML models on the device.	Experimental	AI/ML API reference guide `sl_ml_<model_name>_model_init()` `sl_ml_<model_name>_model_run()`	Standard SiWG917 SoC	Audio Classifier for SiWG917 Voice Control Light for SiWG917
Audio Applications	Audio applications enabled on SiWG917 through software optimizations	Experimental		Standard SiWG917 SoC	Audio Classifier for SiWG917 Voice Control Light for SiWG917

Unsupported Features#

SiWG917 does not use the built-in TensorFlow component directly; instead, support is provided through automatic code generation handled by configurators.
Series 2 devices do not support new software optimizations but, due to their architecture, they are still as efficient as SiWG917, if not more.

SDK Release and Maintenance Policy#

See our SDK Release and Maintenance Policy.