BrainChip announced it is bringing the Akida Neuromorphic System-on-Chip (NSoC) to market, which it describes as a “new breed of neural network acceleration SoCs that puts artificial intelligence at the edge and enterprise” for sectors such as fintech and cybersecurity, among others. Neuromorphic computing is a type of artificial intelligence that is inspired by the biology of the human neuron, also known as spiking neural networks (SNNs). BrainChip’s SNN technology can learn autonomously, evolve and associate information.
SNNs are lower power than traditional convolutional neural networks because they replace the math-intensive convolutions and back-propagation training methods. Each Akida NSoC has 1.2 million neurons and 10 billion synapses, representing 100 times better efficiency than neuromorphic test chips from Intel and IBM. Comparisons to leading CNN accelerator devices show similar performance gains of an order of magnitude better images/second/watt running industry standard benchmarks such as CIFAR-10 with comparable accuracy.
Using the NSoC, users can network many Akida devices together to perform complex neural network training and inferencing. Within the Akida neuron model are training methodologies for supervised and unsupervised training. In the supervised mode, the initial layers of the network train themselves autonomously, while in the final fully-connected layers, labels can be applied, enabling these networks to function as classification networks. The Akida NSoC is designed for either off-chip training in the Akida Development Environment or on-chip training. An on-chip CPU is used to control the configuration of the Akida Neuron Fabric as well as off-chip communication of metadata.