Concept: American technology startup BrainChip and semiconductor startup SiFive have partnered to combine their technologies to offer chip designers optimized AI and ML for edge computing. BrainChip’s Akida technology and SiFive’s multi-core capable RISC-V processors have been combined to create a highly efficient solution for integrated edge AI computation.

Nature of Disruption: With high performance, ultra-low power, and on-chip learning, BrainChip’s Akida is an advanced neural networking processor architecture that takes AI to the edge. SiFive Intelligence solutions combine software and hardware to accelerate AI or ML applications with its highly configurable multi-core, multi-cluster capable design. For AI and ML workloads, SiFive Intelligence-based processors can provide industry-leading performance and efficiency. The highly programmable multi-core, multi-cluster capable design can be used for a range of applications requiring high-throughput, single-thread performance while operating within the most stringent power and area limitations. Akida acts like a human brain, analyzing only the most important sensor inputs at the time of acquisition and processing data with unmatched efficiency, precision, and energy efficiency. BrainChip’s technology is based on its SNAP (spiking neuron adaptive processor) technology, which it licenses to other companies. RISC-V is an open instruction-set computing architecture based on well-known RISC ideas. It provides the high data processing speed that all new and heavier applications require.

Outlook: The duo aims to help companies looking to seamlessly integrate an optimized processor with dedicated ML accelerators, which are required for the demanding requirements of edge AI computing. They plan to use Akida, BrainChip’s specialized, differentiated AI engine, in conjunction with high-performance RISC-V processors like the SiFive Intelligence Series to achieve this. For organizations looking to enter the neuromorphic semiconductor chip market, SNAP provides a development option. It is a key feature of neuromorphic semiconductor circuits that allows for a variety of applications, including cybersecurity, gaming, robotics, and stock market forecasting.

This article was originally published in