Tejasri Gururaj

In a recent study published in Nature Communications, researchers created a memristor that uses a built-in oxygen gradient to produce slow, stable conductance changes, enabling a reinforcement learning (RL) algorithm to learn faster and more stably than conventional approaches.

A new study published in Nature Communications has shown that in the asymptotic limit, extracting the maximum possible work from many copies of a quantum system does not require knowing exactly what state that system is in.

AI data centers are consuming power at rates the US grid was never designed to handle. Natural gas is the only fuel that can be built fast enough to keep up.

By Tejasri Gururaj 40 minutes ago Global data center power demand is projected to hit 84 GW by 2027 —a 50 percent jump from 2023

A new Physical Review Letters study places constraints on the ER = EPR conjecture, showing that under the authors' assumptions, the conjecture would imply possible alterations to the hyperfine structure and effective charge of the hydrogen atom—effects that have never been observed.

A new study published in Physical Review Letters by the IceCube Collaboration reports evidence that the energy spectrum of astrophysical neutrinos is not a simple straight line.

An undisturbed permafrost landscape adjacent to the experimental site on the Tibetan Plateau. Credit: Xiling Gu

A new study suggests a single fiber-optic cable could help detect moonquakes with thousands of sensing points across the lunar surface.

ETH Zurich researchers have developed a chip that cryptographically signs sensor data at the point of capture to fight deepfakes.