Revolutionizing Speed: The Next Generation Optical Memory System for Lightning-Fast Calculations

For many years, there has been a nearly continuous advancement in minimizing the size, while enhancing the performance, of the circuits that energize computers and smartphones. However, Moore’s Law is nearing its conclusion as physical constraints — like the number of transistors that can be accommodated on a chip and the heat generated from densely packing them — are hindering the pace of performance enhancements. The computing capability is slowly leveling off, despite the increasing need for computational power by artificial intelligence, machine learning, and other data-heavy applications. 

Innovative technologies are essential to tackle this issue. A possible answer lies in photonics, which provides lower energy usage and decreased latency compared to electronics.

One of the most encouraging methods is in-memory computing, which necessitates employing photonic memories. Utilizing light signals within these memories allows for executing operations almost instantaneously. However, the proposed methods for developing such memories have encountered difficulties including  slow switching rates and restricted programmability.

Currently, a global team of researchers has established an innovative photonic platform to surmount these challenges. Their research was published in the journal Nature Photonics. 

Collaborating with UC Santa Barbara professor of electrical and computer engineering (ECE) John Bowers and ECE associate professor Galan Moody, project scientist Paolo Pintus, an assistant professor at the University of Cagliari, coordinated the project alongside University of Pittsburgh’s Nathan Youngblood, Institute of Science Tokyo professor Yuya Shoji, and Mario Dumont, who earned his Ph.D. at the Bowers lab.