Research Bits: November 6

Fast superatomic semiconductor; in-memory compute with FeFETs; filtering wireless interference.


Fast superatomic semiconductor

Researchers from Columbia University created a fast and efficient superatomic semiconductor material based on rhenium called Re6Se8Cl2. Rather than scattering when they come into contact with phonons, excitons in Re6Se8Cl2 bind with phonons to create new quasiparticles called acoustic exciton-polarons. Although polarons are found in many materials, those in Re6Se8Cl2 are capable of ballistic, or scatter-free, flow.

In experiments run by the team, acoustic exciton-polarons in Re6Se8Cl2 moved twice as fast as electrons in silicon and crossed several microns of the sample in less than a nanosecond. Given that polarons can last for about 11 nanoseconds, the team thinks the exciton-polarons could cover more than 25 micrometers at a time. And because these quasiparticles are controlled by light rather than an electrical current and gating, processing speeds in theoretical devices have the potential to reach femtoseconds.

Along with working at room temperature, Re6Se8Cl2 can be peeled into atom-thin sheets, a feature that means they can potentially be combined with other similar materials in the search for additional unique properties.[1]

In-memory compute with FeFETs

Researchers from the Technical University of Munich, Robert Bosch, Fraunhofer IPMS, RPTU Kaiserslautern-Landau, and Indian Institute of Technology Kanpur developed an in-memory compute chip that utilizes multi-level ferroelectric field effect transistor (FeFET) cells for multi-bit multiply and accumulate (MAC) operations.

The proposed cell design stores multi-bit information while minimizing device variability effects on accuracy. The FeFETs can store information even when cut off from the power source and use stored data within the memory cell to derive MAC operation results encoded in activation time and accumulated current.

The design achieved 96.6% accuracy for handwriting recognition and 91.5% accuracy for image classification without extra training while delivering 885 TOPS/W. The goal is to use the chip to run deep learning algorithms, recognize objects in space or process data from drones in flight with no time lag, but the researchers expect it will be a few years before this is achieved. [2]

Filtering wireless interference

Researchers from UC San Diego developed a prototype technology to help regulators distribute wireless access at an affordable cost during low-traffic periods. It uses commercial off-the-shelf parts attached to existing radio units with programmable software that allows it to sweep for activity across a range of frequencies within an agency-owned wideband spectrum.

The device can adapt to signal interference in real-time by dynamically adjusting which signals it receives to tune out interference from nearby towers, base stations and other sources of high power signals. The researchers claim the high signal fidelity also enables identification of cyberattacks in real-time. [3]


[1] Jakhangirkhodja A. Tulyagankhodjaev, Petra Shih, Jessica Yu, Jake C. Russell, Daniel G. Chica, Michelle E. Reynoso, Haowen Su, Athena C. Stenor, Xavier Roy, Timothy C. Berkelbach, Milan Delor. Room-temperature wavelike exciton transport in a van der Waals superatomic semiconductor. Science, 2023; 382 (6669): 438

[2] Taha Soliman, Swetaki Chatterjee, Nellie Laleni, Franz Müller, Tobias Kirchner, Norbert Wehn, Thomas Kämpfe, Yogesh Singh Chauhan, Hussam Amrouch. First demonstration of in-memory computing crossbar using multi-level Cell FeFET. Nature Communications, 2023; 14 (1)

[3] Raghav Subbaraman, Kevin Mills, Aaron Schulman, Dinesh Bharadia. Crescendo: Towards Wideband, Real-time, High-Fidelity Spectrum Sensing Systems. ACM MobiCom ’23: Proceedings of the 29th Annual International Conference on Mobile Computing and Networking October 2023 Article No. 79 Pages 1–14

Leave a Reply

(Note: This name will be displayed publicly)