Simulation– tag –
-
Universal MLIPs Face Generalization Challenges, ACS Study Highlights Need for Iterative Fine-Tuning of Material-Specific Models
ACS Publications USA Overview An ACS Publications paper highlights that while Universal Machine Learning Interatomic Potentials (MLIPs) are rapidly becoming general tools for atomic simulations, their role in quantitative material modeli... -
Integrated Active Learning and Knowledge Distillation in MLMD Achieves 1/3 Data Efficiency with MACE Model, Outperforming DeePMD
ACS Publications USA Overview This study developed data-efficient and fast Machine Learning Molecular Dynamics (MLMD) interatomic potentials (MLIPs) by combining DeePMD and MACE models within an active learning and knowledge distillation... -
Machine Learning Interatomic Potentials (MLIPs) Accelerate Catalyst Discovery, Achieving DFT-Level Accuracy at Low Cost
nano-matter.com International Overview As of August 2026, Machine Learning Interatomic Potentials (MLIPs) have become a mainstream technology in computational catalyst research, enabling rapid screening of catalyst candidates, efficient ... -
Mitsubishi Chemical Integrates Generative AI, Quantum Computing, GPU Acceleration for EUV Photoresist Design
SEMICON Taiwan / Mitsubishi Chemical Japan Overview Mitsubishi Chemical is strategically integrating generative AI, quantum computing, and GPU-accelerated computing for next-generation material design, notably for EUV photoresist materia... -
Machine Learning Optimizes Charge Transport Layers and Fabrication Parameters for Lead-Free Cs2AgBi0.75Sb0.25Br6 Perovskite Solar Cells
The Royal Society of Chemistry Unknown Overview This study utilized a machine learning-driven framework to optimize charge transport layer selection and fabrication parameters for eco-friendly, lead-free Cs2AgBi0.75Sb0.25Br6 perovskite s... -
UChicago Lab Discovers Climate-Action Materials via ML-Driven Workflow: Creates Methane-Separating MOFs ‘UCHI-1’ and ‘UCHI-2’
UChicago Pritzker School of Molecular Engineering (PME) USA Overview A research lab at the University of Chicago developed a machine-learning-driven, end-to-end workflow to streamline materials discovery from academic concepts to manufac... -
Novel MLIP Developed for Titanium Carbide MXenes: Applied to Ion Irradiation Simulations, Offering New Defect Engineering Guidance
The Royal Society of Chemistry UK Overview A machine-learned interatomic potential (MLIP) has been developed for titanium carbide MXenes, demonstrating successful application in ion irradiation simulations. Trained with density functiona... -
Mira Proposes “Fifth Paradigm” of AI in Materials Science: Integrating MatterGen, A-Lab, GNoME for Autonomous Discovery
Mira USA Overview Mira has proposed the "Fifth Paradigm" of AI in materials science, asserting that the integration of generative design (MatterGen), quantum chemistry simulations, and closed-loop laboratories (A-Lab) will accelerate aut... -
MACE and SevenNet Data Efficiency Evaluated for Material-Specific MLIP Construction: Achieving Ab Initio Accuracy with 2,000 AIMD Configurations
arXiv International Overview The amount of ab initio molecular dynamics (AIMD) data required to fine-tune universal machine-learned interatomic potentials (MLIPs) for material-specific applications has been quantified. Research indicates... -
LAMMPS Accelerates MLIP Integration: Revolutionizing MD Simulations with DeePMD-kit, MACE, CHGNet Foundation Models
LAMMPS Molecular Dynamics Simulator USA Overview The LAMMPS molecular dynamics simulator is enhancing its integration with several machine-learned interatomic potential (MLIP) frameworks, including DeePMD-kit, NequIP, Allegro, MACE, Seve...
