loading
loading
Off the Wire Press Releases Always-on neuromorphic AI with on-chip learning comes to the world’s most popular single-board computer LAGUNA HILLS, Calif., Oct. 30, 2026 — BrainChip Holdings Ltd., a global leader in ultra-low-power, fully digital, event-based neuromorphic AI, has launched the AKD1500 Raspberry Pi 5 Development Kit. The kit pairs a Raspberry Pi 5 with BrainChip’s AKD1500 M.2 board, giving the millions of engineers, makers and product teams who already build on Raspberry Pi an off-the-shelf way to develop low-power, always-on AI applications – and a clear path to production on the same AKD1500 chip. “As an American Tier 1 for physical AI, Hellbender looks for edge technologies that can move cleanly from prototype to production,” said Brian Beyer, CEO, Hellbender. “BrainChip’s AKD1500 Raspberry Pi 5 kit makes it easy for engineering teams to evaluate low-power, always-on AI on familiar hardware, and we see it as a promising option for the robotics and vision applications our customers are building.” The Challenge: Always-on AI Strains Power, Privacy and the Path to Production A growing class of edge devices – sensors, monitors, wearables, industrial equipment and smart home products – must listen, watch or measure continuously. Running AI on these devices with conventional processors means analyzing every frame and sample, whether anything has changed, which drains batteries and adds heat. Sending the data to the cloud instead adds latency, bandwidth cost and privacy risk. And once a model is deployed, teaching it something new typically means collecting data, retraining in the cloud and pushing an update. Developers face a second hurdle: the hardware used to build a proof of concept is often not the hardware that ships. Moving from a prototype to a production design can mean re-porting models, re-validating performance and starting over on a new platform. The Solution: Event-Based AI That Learns on the Device AKD1500 uses BrainChip’s event-based, neuromorphic architecture to process only meaningful events rather than every piece of incoming data, making it well suited to power always-on workloads such as sensing, monitoring and detection. All processing happens on the device, so applications avoid the latency, bandwidth cost and privacy exposure of the cloud. AKD1500 also supports on-chip learning, so models can learn new classes directly on the device without cloud retraining. A developer can teach a Raspberry Pi 5 application to recognize a new keyword, gesture or object in the field, and the data stays local. “Raspberry Pi is where many edge AI products begin,” said Steve Brightfield, chief product officer, BrainChip. “This kit puts neuromorphic AI where developers already work and gives teams a fast, low-cost way to validate applications on streaming data.” “Kit support will include Docker images that can be preloaded to simplify benchmarking and demos and that serve as a reference for custom neuromorphic application containers. Developers who already have a Raspberry Pi 5 with an M.2 HAT can add the production AKD1500 M.2 card and use the same software.” From Prototype to Production on One Chip The kit is off the shelf and quick to set up. Teams evaluate their application on Raspberry Pi 5, then design the same AKD1500 chip into their own product with a PCIe or SPI interface to suit their system without switching to new silicon. The kit joins the AKD1500 PCIe card and AKD1500 M.2 board in BrainChip’s family of AKD1500 development platforms, and teams can prototype on any of them. For developers, the kit offers: - A familiar platform: Build and test edge AI on Raspberry Pi 5 hardware and workflows developers already know. - Always-on efficiency: Event-based processing handles continuous sensing workloads without processing every input. - Local learning and privacy: Add new keywords, gestures or objects on the device, with no cloud retraining and no data leaving the device. - Free tools and models: BrainChip’s MetaTF software tools and models are available at no charge through the BrainChip Developer Hub. - One chip, prototype to product: The AKD1500 chip that runs on the kit is the same one teams design into production. Availability The AKD1500 Raspberry Pi 5 Development Kit is available now through the BrainChip Shop. MetaTF software tools and models are available at no charge through the BrainChip Developer Hub. About BrainChip Holdings Ltd. BrainChip (ASX: BRN, OTCQX: BRCHF, ADR: BCHPY) is the worldwide leader in edge AI on-chip processing and learning. The company’s first-to-market, fully digital, event-based AI processor, Akida, uses neuromorphic principles to analyze only essential sensor inputs at the point of acquisition, processing data with unmatched efficiency, precision and energy economy. BrainChip’s Sequora is an efficient digital AI processor for GenAI and other large models using state-space architectures. BrainChip’s technology is deployed across industries including aerospace, defense, industrial IoT, healthcare, consumer devices, and wearables. Source: BrainChip
document fetched 2026-10-31 02:54 · canonical https://hpcwire.com/off-the-wire/brainchip-launches-akd1500-raspberry-pi-5-kit