Researchers developed a quantum neural network training framework that enables direct gradient-based optimization on quantum hardware, demonstrating 16-qubit on-hardware training using IonQ's Forte Enterprise system.
Penn physicists demonstrate all-light switching using only 4 quadrillionths of a joule, combining photon speed with electron interaction.
University of Toronto overcomes DNA computing bottleneck with enzymatic synthesis and spatial clustering, enabling months-to-days development timelines.
Ames Lab's agentic AI model identifies new magnet materials without rare earths by coupling fundamental physics with interactive reasoning.
Ames Lab's DuctGPT, a physics-trained model, generates novel magnet compositions without rare earth elements, addressing US supply chain vulnerabilities.
Applied mathematicians find that adding randomness to robot movement prevents gridlock and maximizes swarm efficiency in confined spaces—the Goldilocks zone of controlled chaos.
Monash researchers demonstrate light-based AI chip using quantum valley properties at room temperature, published in Nature Photonics.
The LD-FPG framework generates full-atom protein dynamics and conformational changes, modeling how proteins move and change shape during biological function—not just their static structures.
Applied mathematicians find that adding randomness to robot movement prevents gridlock and maximizes swarm efficiency in confined spaces—the Goldilocks zone of controlled chaos.
DuctGPT, a physics-trained AI developed by Ames Lab researcher Prashant Singh, is designed to discover novel magnet compositions without rare earth elements, addressing US supply chain vulnerabilities.
The LD-FPG framework generates full-atom protein dynamics and conformational changes, modeling how proteins move and change shape during biological function—not just their static structures.
Rentosertib (ISM001-055 / INS018_055), an oral TNIK inhibitor designed using AI, advances to Phase III clinical trials