New superconducting chip aims to help tame fusion plasma
Researchers fed fusion data to a new chip to see whether it could spot signs that superhot plasma is about to go haywire
Scientists at Brookhaven National Laboratory have unveiled a superconducting chip aimed at monitoring fusion plasma in real-time. The chip, developed with the help of the U.S. Department of Energy and AI hardware start-up Great Sky, could help control erratic plasma inside fusion reactors. Fusion reactors use powerful magnets to hold plasma in place, but the plasma can become unstable, causing disruptions that release energy and damage the device.
These disruptions happen rapidly, and analyzing plasma data in real-time can be challenging for traditional computers. Neuromorphic AI systems, designed to mimic the human brain, may be better suited for the task. The superconducting chip, which can process data at a million samples per second, was tested using data from a DIII-D tokamak in San Diego.
The chip successfully flagged anomalies and predicted disruptions without lag. While the chip's capabilities are still being evaluated, researchers hope to develop a larger neural network chip and deploy it at the DIII-D tokamak next year.
Written by urgent.news from Scientific American's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
