COMSOL Multiphysics® version 6.4 introduces new modeling and productivity features, major performance improvements and expanded multiphysics capabilities. New functionality for structural explicit dynamics enables nonlinear dynamic analysis with explicit time integration for high speed events such as impacts, blasts and drop tests, offering robust contact handling for challenging structural simulations.

The new Granular Flow Module, based on the discrete element method (DEM), enables detailed simulation of grain and powder systems in applications such as conveying, mixing and additive manufacturing. Solver performance has been greatly enhanced through the new NVIDIA CUDA® direct sparse solver (cuDSS) for NVIDIA® GPUs, providing several-fold speedups for both single-physics and multiphysics simulations.

Geometry and meshing workflows have been improved with automatic surrounding domain creation, array selections and higher-quality quad-dominant meshing and swept meshing. Visualization updates include spatially varying transparency, customizable environment maps and new array-based results layouts. Usability and productivity have been further advanced with optional large language model (LLM)-assisted simulation; the Chatbot window now supports connections to GPT-5™, DeepSeek™, Google Gemini™, Anthropic Claude™ and other OpenAI API-compatible models, enabling interactive, model-aware assistance that combines the COMSOL documentation with information from an active simulation.

The Application Builder includes several updates for better management of apps with many forms and methods. A new capability in the Model Manager allows batch and cluster studies to use models and data stored directly in a Model Manager database, simplifying version control and shared computing workflows across local and server-based installations.

A summary of the major news in the COMSOL® software version 6.4 is below.

General Updates

  • New Granular Flow Module for simulating the movement and interaction of solid grains in bulk processes
  • NVIDIA® cuDSS for one or more GPUs, providing several-fold speedups for both single-physics and multiphysics simulations
  • Spatially varying transparency
  • Array-based plot layouts
  • Support for OpenAI API-compatible providers such as GPT-5, DeepSeek, Google Gemini, and Anthropic Claude
  • More efficient handling of large simulation apps
  • Cluster support for models saved in a Model Manager database
  • New optimization options for time-dependent and parametric studies
  • Export of network parameters for deep neural network (DNN) surrogate models
  • Surrogate model data generation on clusters
  • Frequency- and time-dependent uncertainty quantification.

Electromagnetics

  • Transient induction boundary conditions
  • Far-field radiation in substrates
  • More efficient switching arc simulation
  • More robust electric discharge simulation
  • Modeling of etching and deposition processes in semiconductor applications
  • Light scattering in tissues and humid environments
  • Simulation of ferroelectric and piezoelectric semiconductors as well as memristors.