Highlights
- BrainChip shares traded unchanged at AUD 0.105 on 16 November, with 910,981 shares changing hands during the ASX session.
- BrainChip introduced an AKD1500 version using a serial peripheral interface to connect with host electronics in edge devices.
- The company uses its Akida architecture to run AI workloads on devices close to where information is collected.
- BrainChip has described development hardware designed to give engineers a practical way to evaluate Akida before developing products.
BrainChip Holdings Limited (ASX:BRN) traded unchanged at AUD 0.105 on 16 November 2026, with shares moving between AUD 0.100 and AUD 0.108. The company has been expanding how engineers can work with its Akida architecture. Its AKD1500 chip with a serial peripheral interface brings the focus to everyday edge devices and how developers connect AI processing to existing electronics.
Processing Takes Place Close to the Input
Edge processing runs work on or close to the device collecting the information. In BrainChip’s case, the company uses its Akida architecture to carry out AI inference. Inference is the stage when an existing model processes new input and produces an output, such as a classification.
The architecture is intended for devices where power and other hardware constraints influence how a design can work. BrainChip’s chip release describes a path to adding this processing locally to products where a developer needs to evaluate information gathered by the device.
The Interface Gives Engineers a Connection Point
The AKD1500 version announced by BrainChip uses a serial peripheral interface, commonly shortened to SPI. That interface provides a route for a host system to communicate with the processor. The interface matters to the developer because the chip must work with the host electronics as part of a larger design.
A developer evaluating the chip also needs to consider what data the application collects, how it will send that data to the processor and how it will use the result. BrainChip pitches its hardware for edge use cases where those steps need to happen close to the sensor.
Development Hardware Gives Teams a Place to Test
BrainChip has also outlined an AKD1500 development kit for Raspberry Pi 5. The company positions the kit as a way for engineers to test the processor in a familiar development environment and work toward a finished design.
Evaluation and production involve different decisions. A kit allows a team to run experiments before settling on the components it will use in a product. The chip’s interface then becomes relevant when engineers decide how to connect local AI processing to the rest of the electronics. BrainChip’s hardware releases address these different stages.
The kit pairs a Raspberry Pi 5 with BrainChip’s AKD1500 M.2 board. BrainChip also describes on-chip learning, which lets developers test recognition of additional classes, such as a new keyword or gesture, using the device itself. This gives teams a way to evaluate both inference and learning on development hardware.






Please wait processing your request...