BrainChip, Neuromorphyx launch AKD1500 developer board for embedded AI

BrainChip and Neuromorphyx announced BrainBoard1500, a compact developer board built around BrainChip’s AKD1500 neural-network accelerator. The platform aims to make neuromorphic AI easier for embedded engineers to test, measure and adapt for low-power edge devices.

Why it matters: - BrainBoard1500 gives embedded engineers a smaller, more direct way to evaluate BrainChip’s AKD1500 for edge devices with tight power, latency and size limits. - The board is aimed at developers who want to test spiking neural network models on sensor data without moving to larger M.2- or PCIe-based lab hardware. - The launch broadens access to neuromorphic AI silicon for education, industrial and embedded AI use cases.

What happened: - BrainChip and Neuromorphyx announced the BrainBoard1500, a compact developer and evaluation board built around BrainChip’s AKD1500 neural-network accelerator. - Neuromorphyx designed and manufactured the board as part of a partnership with BrainChip. - The board is positioned for the embedded developer community and supports the Arduino Nicla form factor. - The companies said the initial out-of-the-box software release is supported through the Neuromorphyx Arduino library and drivers.

The details: - BrainBoard1500 includes direct connector compatibility with Arduino Nicla boards. - The board ships with open-source Arduino Nicla library and drivers. - It uses a direct SPI/QSPI path for running spiking neural network models on sensor data. - The board also includes I2C system control and onboard model flash. - A two-channel current monitor supports real-time power profiling of AKD1500 inference workloads. - The board can be used to evaluate sleep and wake configurations. - The standard SPI/QSPI architecture is intended to support software adaptation across ecosystems including Raspberry Pi, Seeed Studio/XIAO, OpenMV, Espressif, DFRobot, Adafruit, SparkFun, STMicroelectronics and custom MCU and FPGA hosts. - Neuromorphyx created the hardware architecture, electronics, PCB manufacturing design and Arduino-oriented software path, including the board library, drivers and Akida Engine integration. - BrainChip supplies AKD1500 chips and technical enablement packages. - BrainChip will support customer-ready AI model repositories for keyword spotting, visual wake words and human activity recognition. - BrainBoard1500 will initially be available through the Neuromorphyx store, with a full product page and supporting materials. - The package will include an open-source Arduino software repository, datasheet and supporting documentation. - BrainChip and Neuromorphyx plan broader distribution through education and approved ecosystem channels.

Between the lines: - The product appears designed to remove a common barrier in edge AI development: the gap between lab evaluation boards and the hardware constraints of compact devices. - The partnership also gives BrainChip a more approachable developer entry point for its Akida platform. - Neuromorphyx is extending its existing neuromorphic vision work into a smaller board that fits more easily into maker and embedded workflows.

What's next: - BrainBoard1500 will launch first through Neuromorphyx’s store. - The companies plan to expand availability through additional distribution, education and approved ecosystem channels. - BrainChip is expected to continue adding model repositories for common embedded AI tasks. - Developers can use the board to run, measure and refine spiking neural network models on the AKD1500 device.

The bottom line: - BrainChip and Neuromorphyx are trying to make neuromorphic AI hardware easier to test, teach and deploy at the edge.