Matlantis Announces Major Upgrade to Its Universal Atomistic Simulator for Materials Discovery, Opens Dedicated U.S. Office

Article content

Training datasets for Matlantis’s core AI technology are now developed using r²SCAN, doubling accuracy in atomistic simulations compared to previous version

Article content

CAMBRIDGE, Mass., July 16, 2025 (GLOBE NEWSWIRE) — Matlantis Inc., the U.S. hub of the materials‑discovery arm of Japan’s leading AI company Preferred Networks, Inc. (PFN), today announced a major update to its Matlantis™ universal atomistic simulator, and the opening of its office in Cambridge, Massachusetts for accelerating adoption of AI‑driven materials research across North America. The update introduces the new Version 8 of PFN’s proprietary AI technology named PFP (Preferred Potential), which enables researchers across industries to accelerate discovery, improve predictive performance, and unlock new frontiers in materials science with unprecedented levels of simulation accuracy.

Article content

Article content

Article content

PFP Version 8 marks a significant milestone as the first universal machine learning interatomic potential (MLIP) to be trained with datasets developed with a new method called r²SCAN (restored-regularized strongly constrained and appropriately normed) functional. PFP versions up to 7 relied on datasets generated with a method called PBE (Perdew-Burke-Ernzerhof) functional, which has also been widely adopted by MLIPs other than PFP. It is known, however, that PBE has certain limitations in simulation accuracy—how closely computer-based simulations of materials’ behavior align with real-world experimental results.

Article content

Article content

The introduction of the r²SCAN method is the culmination of PFN’s continuous efforts over the past couple of years to overcome the accuracy limitations of the PBE-based approach. Developing training datasets with the r²SCAN method is more computationally intensive, requiring three to five times the computing time compared to the PBE method. However, because PFP Version 8 is now trained with the datasets built with r²SCAN as well as PBE, Matlantis users can achieve up to doubled the simulation accuracy in the same timeframe as the previous version.

Article content

Article content

“This update represents a significant breakthrough,” said Daisuke Okanohara, CEO of Matlantis. “In 2021, we were the world’s first to launch a commercial simulator using a universal MLIP, and now our simulator, Matlantis, is the first globally to incorporate r²SCAN that ensures high simulation accuracy. We believe this will further pave the way for the era of computer-based materials discovery. We will continue to support researchers in North America and the rest of the world to discover innovative and sustainable new materials.”

Article content

Jointly invested by PFN and ENEOS, Japan’s largest energy company, and Mitsubishi Corporation, Matlantis has already been used by over 100 industrial and academic leaders worldwide since its launch in July 2021. Today, Matlantis is among the first commercially available AI‑powered platforms purpose‑built for atomistic simulation at industrial scale—offering a single MLIP that spans 96 elements (from hydrogen to curium) and delivers DFT (density functional theory)‑level accuracy up to 20 million times faster.

Article content

Matlantis enables research teams to: