The evaluation board has been developed to complement the Company’s Akida Development Environment (ADE) for Early Access Partners (EAP) that have developed Akida compatible networks. ADE is a development environment that allows the design of a neural network as a Convolutional Neural Network (CNN) and. The ADE workflow can also be used to convert the network to an Event-Based CNN or develop a native Spiking Neural Network (SNN).
Implementing Event-Based CNNs significantly reduces power consumption by processing only non-zero activations, taking advantage of the sparsity in most data. Additional power savings can be captured by aggressive quantization without a meaningful loss of accuracy for the majority of network designs. Once converted to an Event-Based CNN network, both weights and activations can be quantized to 1, 2 or 4 bits selectively.
The Akida neural fabric also offers new features that provide true AI Edge Learning rather than AI Edge Training. Training requires connectivity through the cloud or a data center to re-train the network for new classifiers. The ability to implement AI Edge Incremental Learning is a new and valuable feature in AI Edge products.
BrainChip CEO, Louis DiNardo commented, “The Akida NSoC has proven to provide significant power savings and is the complete integration of a neural network design. Akida introduces new and powerful features to the high-growth AI Edge market.”
“True AI Edge learning does not exist with current AI solutions and our ability to provide this and other features while significantly reducing both power consumption, size and the Bill-of-Materials (BOM) has attracted the attention of leading suppliers in the Smart Home, Smart Transportation and Smart City markets," says DiNardo. "We believe that these capabilities also will provide new solutions in Smart Healthcare applications that serve our mission of supporting Beneficial AI applications and improve the human condition globally in terms of diagnosis of infectious diseases, cancers and a wide array of challenging global health concerns. We are excited about our technology and the potential to impact many industries including healthcare and energy conservation which are clearly a global concern.”
The Akida NSoC is high-performance, small, ultra-low power and enables a wide array of edge capabilities.