SCM Releases ‘AMS2026’ Software, Accelerating Materials Chemistry Simulations with ML Potentials and GPU Optimization

SCM (Scientific Computing & Modelling) Netherlands

Overview

SCM announced the release of ‘AMS2026’ software, featuring major advancements in machine learning potentials (eSEN, MACE, UMA) that expand chemical coverage for biomolecules, catalysts, MOFs, and inorganic materials. This update delivers GPU-optimized performance, enhanced electronic structure functions, improved OLED and multiscale workflows, smart GUI tools, and VASP integration. AMS2026 streamlines automated, reproducible research, significantly boosting materials simulation productivity.

In Depth

Key Findings

Scientific Computing & Modelling (SCM) has released ‘AMS2026,’ the latest version of its comprehensive simulation software designed for researchers in materials science and chemistry. This version is characterized by significant advancements in machine learning potentials (eSEN, MACE, UMA), which substantially broaden the chemical coverage for diverse material systems such as biomolecules, catalysts, metal-organic frameworks (MOFs), and inorganic materials. The software features GPU-optimized performance, enabling faster workflows and the derivation of more accurate results.

Technical / Clinical Details

AMS2026 simultaneously enhances the accuracy and efficiency of materials simulations by integrating traditional physics-based computational methods with cutting-edge machine learning techniques. Specifically, the new machine learning potentials—eSEN, MACE, and UMA—allow for a faster and more precise description of interatomic interactions, making large-scale molecular dynamics and Monte Carlo simulations feasible. GPU-optimized computational kernels dramatically accelerate the execution speed of these simulations, helping researchers solve complex problems in less time. Furthermore, enhanced electronic structure calculation capabilities improve the accuracy of predicting material electrical and optical properties, contributing to the design of Organic Light-Emitting Diode (OLED) devices and advancements in multiscale modeling workflows. The inclusion of smart Graphical User Interface (GUI) tools and integration with the widely used VASP (Vienna Ab initio Simulation Package) are designed to enable researchers to conduct computational studies more easily and reproducibly.

Background & Context

The discovery and development of new materials are driving innovation across many industries, including energy, electronics, and pharmaceuticals. However, the process of predicting material behavior at the atomic level and optimizing its functionality is a complex challenge requiring vast computational resources and time. The advancements in machine learning, particularly in machine learning potentials, have emerged as powerful means to address this challenge. They allow for the simulation of large-scale systems while significantly reducing the cost of high-precision calculations based on quantum mechanics. The release of AMS2026 provides the latest research findings in this field to industry and academia, supporting the acceleration of materials development through the fusion of AI and high-performance computing.

Strategic Significance & Outlook

AMS2026 paves the way for materials scientists to explore more complex and realistic material systems. It will enable deeper insights and faster design cycles in areas such as biomolecule-material interactions, advanced catalytic reactions, the design of novel MOFs, and the development of next-generation inorganic materials. The utilization of GPUs will become increasingly indispensable for even larger simulations and AI model training in the future, driving the overall development of the materials informatics field. SCM is expected to continue providing automated and reliable computational science tools through ongoing software development, addressing new challenges faced by researchers.