In the face of an ever-growing demand for computational power, spurred on by the technological revolution, scientists are tirelessly working to enhance the existing processing capabilities and pioneer novel computing methodologies. Two recent studies by the research team led by Jean Anne Incorvia, a professor at the Chandra Family of Electrical and Computer Engineering in the Cockrell School of Engineering, are making significant strides in both these directions. They are not only improving the current semiconductor technology but also paving the way for a new breed of computers that mimic human brain functions.
According to Incorvia, “We stand at the threshold of a new era of computing and replicating how our brains work is a massive research challenge. However, we cannot overlook the importance of enhancing and innovating the devices that power our existing technology as they continue to be relevant.”
The first study, published in ACS Nano, focuses on transistors and circuits. Chips contain components called logic gates that process digital signals. These gates are transistors that can typically conduct either electrons or holes but not both. Holes are created when electrons move within atoms. In this study, the team managed to connect logic gates capable of conducting both electrons and holes, thereby reducing the number of transistors required in a circuit. This allows for more transistors to be accommodated within the same space, enhancing efficiency and power, or alternatively, the saved space could be used to miniaturize the device.
The transistors were made from ultra-thin two-dimensional materials with an inherent “ambi-polar” property, enabling them to conduct both holes and electrons. However, they were initially not very efficient at it. The researchers refined this capability and demonstrated important XOR, NOR, and NAND circuits using only these ambi-polar transistors. These circuits form the fundamental building blocks of larger circuits. Incorvia believes that if they can leverage the natural behavior of these 2D materials and scale them, they could potentially halve the number of transistors needed in their circuits.
The second study, published in Applied Physics Letters, delves into the future of computing: neuromorphic devices that function more like the human brain. These devices excel at AI tasks such as image interpretation and language processing compared to traditional computers. The researchers developed a new type of artificial neuron using magnetic materials. These artificial neurons are unique due to their chaotic response to electric pulses, allowing them to outperform other artificial neurons in interpreting images, especially when dealing with noisy data sets.
These artificial neurons could be revolutionary for “edge computing” applications, where devices need to be smaller, consume less power, and operate far from a central computing source like a cloud server. They also exhibit resistance to radiation, making them ideal for potential use in outer space where silicon chips struggle against high radiation levels.
Incorvia collaborated with fellow faculty members Deji Akinwande and Joseph Friedman from electrical and computer engineering on the ACS Nano research. The project was funded by grants from the National Science Foundation and U.S. Air Force Research Laboratory. The Applied Physics Letters research was also funded by grants from the National Science Foundation.
These groundbreaking studies mark significant advancements in electronics and computer engineering. They not only enhance our understanding of semiconductors but also lay the groundwork for developing computers that function like the human brain. This research has far-reaching implications for programming languages and coding practices in the future. The ability to create more efficient circuits and artificial neurons could revolutionize electronics industry and pave the way for new products that cater to our ever-increasing computational needs.
