June 3's joint disclosure said Quemix and Honda R&D built the first quantum algorithm designed to accelerate Density Functional Theory calculations, opening a path to simulations of extremely large-scale materials systems beyond conventional computers.
Demonstration tests on an emulator showed calculation time falling exponentially as problem size grew, while maintaining accuracy comparable to conventional DFT in key material-property calculations.
The advance works by bypassing DFT's nonlinear Gram-Schmidt orthogonalization step and directly computing total energy with Quantum Phase Estimation circuits, avoiding costly electron-density readouts.
Honda R&D said the speed-focused approach targets weakly correlated materials common in semiconductors, batteries and drug discovery, where existing DFT is accurate enough but too slow for broad industrial use.
Both companies now plan to run the algorithm on actual quantum hardware, aiming to shorten materials-development cycles and push broader adoption of computational Materials DX.
Will this quantum algorithm compete with or complement AI in the race to discover new materials?
Does this quantum speedup risk generating wrong answers faster, or does it also improve the reliability of material simulations?
Honda promises exponential speed, but will this breakthrough remain theoretical until fault-tolerant quantum computers arrive years from now?
Quantum Leap in DFT: Exponential Acceleration by Quemix and Honda R&D Reshapes Materials Discovery
Overview
Quemix Inc. and Honda R&D have announced a major breakthrough by developing the world’s first quantum algorithm that can exponentially speed up Density Functional Theory (DFT) calculations. DFT is essential for predicting material properties at the atomic level and is a key part of the Materials Digital Transformation. Traditionally, DFT’s heavy computational demands have limited the size and complexity of simulations. The new method uses Quantum Phase Estimation circuits to directly calculate total energy, bypassing the need for explicit electron density readout. This innovation overcomes long-standing computational barriers and opens new possibilities for materials science.