Computing Gets a Boost from New Research

Two new papers published by the Cockrell School of Engineering’s Chandra Family of Electrical and Computer Engineering provide blueprints for improvements in semiconductor technology and building blocks for the next generation of computers that mimic the human brain. Published in ACS Nano, the research relates to logic gates that interpret digital signals.

Logic gates are transistors that conduct either electrons or holes, but not both. In the paper, the researchers linked logic gates that could conduct both, reducing the number of necessary transistors. Engineers could then pack additional transistors into that space or miniaturize the device.

Refining the capability of transistors to conduct both holes and electrons is a major component of this paper. Through their device engineering, they show important XOR, NOR, and NAND circuits without any other devices but the ambipolar transistors. The researchers proved the capability with a large device size. The next steps are to shrink the apparatus and reduce the power consumption needed to make them commercially viable.

A second paper published recently in Applied Physics Letters looks at the next generation of computers that think more like the human brain. These devices are better at AI tasks like interpreting images and language processing. Researchers created a new type of artificial neuron using magnetic materials.

These neuron devices have a chaotic nature in their reactions to electric pulses. They outperformed other artificial neurons as part of neural networks in interpreting images, specifically when the data analyzed was noisy. The neurons could be important for “edge computing” uses, where devices are smaller, use less power, and are far removed from a central computing source like a cloud server. They are also resistant to radiation. In fact, one of the initial applications of this technology could be space, where silicon chips find it difficult to stand up to the high level of radiation.