What is it about?
In today’s computers, processors that fetch data are slow and consume extensive power. To build faster and better computers, scientists are modeling processors on the human brain. A recent review article talked about these “neuromorphic” processors. In the brain, information moves between cells via spikes of electrical activity. Neuromorphic processors aim to replicate this using “spiking neural networks.” The paper talked about how such “spiking” neuromorphic processor designs have evolved. Then, the review article explains about the hardware used in neuromorphic processors and its properties. The human brain is an “analog” system (based on continuous electrical activity within its cells). The components in processors mimic their biologic counterparts to work as an analog circuit. Other systems use a mixed approach (analog-digital). Studying these can aid the development of neuromorphic processors.
Featured Image
Photo by Josh Riemer on Unsplash
Why is it important?
Neuromorphic networks present a world of opportunity. They can help us develop more efficient computers for the future. Neuromorphic processors, too, are very adaptable. They can be reconfigured and adapted based on the specific job they need to do. Hence, neuromorphic systems have a large variety of applications. The study of neuromorphic systems can help further other fields too. They can help us better understand how the brain functions. KEY TAKEAWAY: Neuromorphic systems have advanced since the development of the first artificial brain network. Studying them will help us build better computers in the future.
Read the Original
This page is a summary of: A system design perspective on neuromorphic computer processors, Neuromorphic Computing and Engineering, November 2021, Institute of Physics Publishing,
DOI: 10.1088/2634-4386/ac24f5.
You can read the full text:
Contributors
Be the first to contribute to this page