BrainChip partners with Edge Impulse for a platform that mimics the brain

BrainChip and Edge Impulse have announced a partnership to deliver next-gen platforms for machine learning. The companies are aiming to make BrainChip’s neuromorphic technology based on spiking neural networks mainstream.

Both companies will combine their technologies to realize faster development cycles, improve time to market and achieve a competitive advantage. This should ultimately foster adoption of neuromorphic computing, which, roughly speaking, mimics the brain. 

BrainChip, as the world’s first commercial producer of neuromorphic IP, will contribute its ultra-low power Akida neuromorphic edge AI processor architecture. Edge Impulse is a leading company for edge AI development that delivers platforms, tools and services for sensor, audio and vision applications.

Edge Impulse and BrainChip are addressing the goal of developers to create smart devices by implementing AI/ML capabilities into their products. “By integrating solutions, such as deploying BrainChip’s neuromorphic IP with our ML platform, developers and enterprise customers are empowered to build advanced machine learning solutions quickly and efficiently so that they are well-positioned as leaders within their respective markets,” said  Zach Shelby, cofounder and chief executive of Edge Impulse.

The partnership combines BrainChip’s unique IP with Edge Impulse’ design platform into one solution. “We continue to look for partnerships that will advance opportunities to enable a wide array of edge AI capabilities, such as continuous learning and real-time inference. Edge Impulse is a perfect addition to our ecosystem of providers enabling next-generation technology to markets around the world,” said BrainChip chief marketing officer Jerome Nadel.

Neuromorphic computing works based on asynchronous spiking neural networks, similar to the brain. This architecture supports on-chip learning and learning based on a little training data. It has been shown to offer large improvements in performance and power in select applications. Although the technology has been researched by various organizations, it has only been commercialized by BrainChip in January of this year. In April, BrainChip also announced a partnership with SiFive.