Edge AI chip provider BrainChip has announced the release of its AKD1500 chip, built on the Akida neuromorphic processor, now available in a compact M.2 module form factor. This new product offers a fanless, plug-and-play solution for edge AI in industrial and commercial designs.
Edge AI chips are specialized processors designed to run artificial intelligence algorithms locally on a device, rather than relying on cloud servers. The new module conforms to the M.2 2230 standard, one of the smallest available sizes, and features a B+M key interface. Its design allows hardware engineers to add on-device AI to existing systems with minimal cost, avoiding the need to redesign power or thermal management solutions. This effectively transforms a hardware upgrade that could take multiple quarters into a simple plug-and-play process.
The AKD1500 is built on a purely digital, event-driven neuromorphic architecture. It processes data only in response to "spikes" or event triggers, avoiding the always-on computational energy consumption of traditional AI chips. In terms of performance, the chip delivers near-tera-operations-per-second efficiency while consuming less than 300 milliwatts in PCIe mode and under 200 milliwatts in serial mode. This low power profile enables integration into battery-powered industrial sensors and portable devices without the need for active cooling fans.
According to BrainChip's Chief Product Officer, power consumption and cost have been the two biggest barriers to bringing AI into fanless industrial equipment and battery-operated commercial devices. The M.2 module was specifically created to address these obstacles, allowing engineering teams to add AI capabilities to existing designs without altering their power or thermal configurations. The module is now available for purchase through BrainChip's official online store.
Comments