Recently published in the scientific journal Nature, researchers at the University of Limerick (UL) have discovered new methods to probe, control, and tailor materials at the molecular level.
The research team, led by Professor Damien Thompson, director of SSPC, the Research Ireland Centre for Pharmaceuticals, collaborated with scientists from the Indian Institute of Science (IISc) and Texas A&M University on this international project. Their work has resulted in a novel hardware platform for artificial intelligence that offers significant improvements in computational speed and energy efficiency.
Professor Thompson explained the innovative approach: "The design draws inspiration from the human brain, using the natural wiggling and jiggling of atoms to process and store information. As the molecules pivot and bounce around their crystal lattice, they create a multitude of individual memory states."
This molecular-level computing method allows for information processing and storage on an incredibly small scale, with each "memory entry" being smaller than an atom. The researchers believe this could lead to massive improvements in energy efficiency and space economy for various computing applications, from data centres to digital maps and online gaming.
The team's reconceptualisation of computing architecture has achieved high-resolution performance, a longstanding challenge in neuromorphic computing. Their system can perform resource-intensive workloads with an unprecedented energy efficiency of 4.1 tera-operations per second per watt (TOPS/W).
Professor Sreetosh Goswami, the project lead at IISc, highlighted a significant achievement: "We created the most accurate, 14-bit, fully functional neuromorphic accelerator integrated into a circuit board that can handle signal processing, AI and machine learning work.’’
This breakthrough could extend neuromorphic computing beyond niche applications, potentially unleashing the transformative benefits of artificial intelligence across various sectors.
The research team is now working on expanding the range of materials and processes used to create these platforms and further increase processing power. Professor Thompson envisions a future where traditional computers are replaced by high-performance "everyware" – energy-efficient and eco-friendly materials providing ubiquitous information processing integrated into everyday items.