Before we move on to the direct news in COMSOL Multiphysics 6.4, Ed Fontes points out the company’s Application Builder as an important strength, and a solution that has played a significant role in the spread of COMSOL’s products to a total of nearly 200,000 users globally.
The point of the Application Builder is to create custom simulation apps based on COMSOL Multiphysics models. To convert these apps into standalone, compiled programs (executable files, eg .exe), the add-on product COMSOL Compiler is required. With this, you can:
• Generate a standalone executable file that can be distributed to anyone.
• These compiled apps can be run without the user having a COMSOL Multiphysics or COMSOL Server license installed.
• The compiled apps can be distributed for Windows, macOS, and Linux operating systems (although the apps must be built in the Windows version of Application Builder).
“Application Builder has spread the Value of Simulations Beyond just Expert Level”
In summary, Application Builder is the design tool, and COMSOL Compiler is the tool that enables compilation into standalone apps. With these, you can spread simulations to a much wider circle of users, coworkers, both internal and external, and to other stakeholders in the development teams. A solution that has thus contributed to the democratization of simulation and analysis work, and has become a major commercial success for COMSOL. Ed Fontes again:
“So it is, and in terms of efficiency and success in product development processes, it is crucial that models can be placed in the hands of those who best understand the processes, components, devices, or machines being developed. Modeling experts can build apps, while product experts use these apps to develop the actual products. This is a highly efficient way to make use of modeling and simulation, especially in rapidly evolving areas,” says COMSOL’s VP of Development.
That said, there are also, as noted in the introduction, connections to the developed NVIDIA CUDA support in the new 6.4 version: It also accelerates COMSOL’s compiled simulation apps.
Faster Simulations with GPU-Accelerated Solvers
Overall, the new COMSOL Multiphysics, version 6.4, introduces new functionality, major performance improvements, and expanded capabilities for multiphysics modeling and simulation app development. As noted in the introduction, the greatly improved solver performance through NVIDIA cuDSS, based on the electronics giant’s CUDA platform for NVIDIA AI infrastructure, is an interesting new feature. With this GPU accelerator for direct solvers, for example, comes expanded capacity for acoustic simulations (see image above) with multiple GPUs.
“The solver also provides significant performance gains even on relatively inexpensive NVIDIA GPUs,” adds Ed Fontes.
Generally, the significant expansion of GPU acceleration support, was something that had previously only been available in some niche workflows within the solution.
We are thus talking about an important step forward in COMSOL’s continued work to improve solver performance and scalability.
In the press material, the company writes that, ”cuDSS, a GPU-accelerated sparse direct solver optimized for hybrid CPU–GPU computation, supports all recent NVIDIA GPU architectures. Depending on the hardware and model characteristics, cuDSS can provide substantial speedups compared to CPU-based direct solvers. GPU acceleration is beneficial for both single-physics and multiphysics simulations, particularly in cases where solver robustness is important. In benchmarks, some multiphysics simulations have achieved speedups of 5x or greater.
”The integration of cuDSS into COMSOL Multiphysics marks a critical step in bringing accelerated computing to the heart of engineering simulation,” said Tim Costa (pictured to the left), general manager of industrial engineering at NVIDIA. “Engineers can now explore larger design space with greater fidelity, reshaping how industries design, validate, and optimize the products that power our world.”
Additionally, NVIDIA CUDA-X cuBLAS library is accelerating the GPU formulation for transient pressure acoustics simulations which can now be run on multiple GPUs on the same machine, or even on a GPU cluster. These improvements reduce computation time significantly for larger models.
Simulation of Motion and Interaction Between Solid Particles in Bulk Processes
But Fontes points out several favorites among the news in 6.4:
”Yes, the new Granular Flow Module another great example. This module is really exciting. Partly because it introduces a new technology for us – DEM (the Discrete Element Method) – and partly because it enables customers to solve many interesting problems, for example in the food and process industries, as well as in the production of powders and ceramics in the battery industry.
The Granular Flow Module, a new add-on product based on the discrete element method (DEM), enables engineers and researchers to model granular processes such as hopper discharge, silo storage, chute transport, powder spreading, and mixing. The Granular Flow Module applies to a variety of industries including pharmaceuticals, chemical processing, agriculture, mining, and additive manufacturing, among others.
By capturing particle-scale effects such as collisions, adhesion, and rotational resistance, and providing detailed control over grain properties, release conditions, and wall interactions, the module helps users evaluate flow uniformity, packing density, mixing efficiency, and wall stresses — revealing issues like blockages or uneven flow to support better process design and optimization.
Expanded AI Support in the Chatbot
That said, let’s move on to Ed Fonte’s third favorite improvement to Multiphysics 6.4, which relates to the AI side.
”Yes, in addition to the major updates, geometry, meshing, and visualization workflows in COMSOL Multiphysics the solution has been en enhanced, and productivity has been further advanced through optional large language model-assisted simulation, (LLM): ”The expanded chatbot window with support for GPT-5, DeepSeek, Google Gemini, Anthropic Claude, and other OpenAI-compatible language models. This is especially important for users who build advanced apps that involve extensive programming with Model Methods to automate sequences of operations. It saves an enormous amount of time for users who do not have significant programming experience,” notes Fontes.
New Possibilities for Time-Explicit Dynamic Analysis
Version 6.4 also introduces a new framework for time-explicit dynamic analysis.
”Yes, it’s a great solution,” comments Ed Fontes, explaining that it, ”enables efficient simulation of fast, transient, and highly nonlinear events such as impact, crushing, and elastic wave propagation.”
The explicit formulation supports a wide range of nonlinear structural materials, including hyperelastic, plastic, viscoplastic, and creep models, and can also be combined with dynamic fracture simulations. To streamline model setup for complex mechanical assemblies, new functionality automatically detects and defines contact conditions between interacting parts.
News Across the Product Suite
Additional highlights of COMSOL Multiphysics version 6.4 include:
• Higher-quality quad-dominant meshing and swept meshing
• Spatially varying transparency
• Array-based plot layouts
• More efficient building of large simulation apps
• 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
• Import of CFD data in CGNS format
• Frequency- and time-dependent uncertainty quantification.
