Researchers from University of Southern California Viterbi School of Engineering Electrical and Computer Engineering have designed a new analog integrated circuit and architecture that can extend where these semiconductors are used beyond traditional applications and into cutting-edge sectors like neural networks.
This technology could also extend to other types of memory technologies as well, such as magnetic memories that use the same device as read head of the magnetic hard disk drives and phase change memories that use the same material as CDs.
Researchers developed the IC architecture and corresponding algorithm to program an analog device precisely to a target value. The analog chip has a higher efficiency — about 10x — and higher speed.
The innovation could be used to train neural networks which are needed to develop artificial intelligence (AI) and machine learning (ML). This is something that could only been done previously with very expensive digital systems.
The analog IC and architecture could also enable new applications beyond AI and ML such as scientific computing for weather forecasting.
The full research can be found in the journal Science.
