Neuromorphic Computing Delivers Low-Power AI
What you’ll learn:
- What are spiking neural networks?
- What is neuromorphic computing?
- Where does BrainChip’s Akida fit in?
Spiking neural networks (SNNs) are an implementation of neuromorphic computing, an aspect of artificial intelligence and machine learning (AI/ML). Neuromorphic computing emulates the operation of physical neurons like those found in the human brain.
SNNs have a similar layered model connectivity as other artificial neural networks (ANNs). However, SNNs use an event- or pulse-based distribution of information through the network whereas ANNs exploit wave-based distribution. As a result, SNNs are more power-efficient.
In this episode of Inside Electronics, Electronic Design’s Senior Content Director, Bill Wong, talks with Steven Brightfield, Chief Marketing Officer at BrainChip, about SNNs and the company’s Akida platform.
Akida is available as IP for inclusion in custom chips and is licensed by companies like Renesas. It’s also available from BrainChip (see figure). An M.2 version is in the works.