New Computer Memory Technology Can Power Future Artificial Intelligence

2023-06-27

A research team led by the University of Cambridge has developed a novel computer memory design that promises to significantly improve performance while reducing the energy requirements of the internet and communication technologies. According to the university, artificial intelligence, algorithms, internet usage, and other data-driven technologies are estimated to account for over 30% of global electricity consumption in the next decade. Dr. Marcus Heronbrand, the lead author from the Department of Materials Science and Metallurgy at the University of Cambridge, stated, "To a large extent, this surge in energy demand is due to the inadequacy of existing computer memory technologies. In traditional computing, there is a back-and-forth transfer of data between memory and processing, which consumes both energy and time." The researchers conducted experiments on a new technology called resistive switching memory. Unlike traditional storage devices that encode data into two states (one or zero), this new type of memory can achieve a continuous range of states. This is achieved by applying current to a specific material, causing an increase or decrease in resistance. The different changes in resistance can create different states to store data. Dr. Heronbrand explained, "For example, a typical USB storage device based on a continuous range can accommodate 10 to 100 times more information." The team developed a prototype device based on hafnium oxide, which has been a challenging issue in resistive switching memory applications due to its lack of atomic-level structure. However, Heronbrand and his colleagues found a solution by introducing barium into it. When barium is added, it forms highly structured barium "bridges" between thicker hafnium oxide films. At the points where these bridges meet the device, an energy barrier is formed, allowing electrons to pass through. The energy barrier can be raised or lowered, thereby changing the resistance of the hafnium oxide composite material and creating multiple states within the material. "These materials are exciting because they can work like synapses in the brain: they can store and process information in the same location, just like our brains," said Heronbrand. The researchers believe that this could lead to the development of computer memory devices with higher density and performance but lower energy consumption, making it particularly promising in the fields of artificial intelligence and machine learning. The technology has been patented through Cambridge Enterprise, the commercialization arm of the University of Cambridge, and scientists are now collaborating with the industry for larger-scale feasibility studies. They claim that integrating hafnium oxide into existing manufacturing processes will not pose difficulties, as this material is already used in semiconductor production. Source: https://thenextweb.com/news/new-computer-memory-tech-could-power-ai-of-future

RECENT AI NEWS

RECENT AI TOOLS