SimScale, the world’s first AI-native cloud engineering simulation platform, announced a strategic collaboration with AI Engineering GmbH to integrate the PAMICS® solver into the SimScale ecosystem. Leveraging accelerated computing on NVIDIA AI infrastructure, the integration removes meshing bottlenecks and dramatically reduces simulation runtimes for complex industrial applications that have traditionally struggled with grid-based methods, delivering simulation speeds 10-20x faster.
By combining AI Engineering’s state-of-the-art Smoothed Particle Hydrodynamics (SPH) solver with SimScale’s cloud-native infrastructure, this partnership aims to democratize access to high-fidelity, meshless Computational Fluid Dynamics (CFD).
This performance leap further positions SimScale as a high-velocity source of synthetic physics data, enabling teams to generate the scale and fidelity required for Physics AI model training and predictive Digital Twins. By combining cloud-native simulation, centralized data management, and accelerated computing, SimScale helps lay the foundation to train the next generation of AI models. This foundation supports downstream Physics AI workflows across the NVIDIA ecosystem, including physics-informed models built directly in SimScale with NVIDIA PhysicsNeMo.
The integration also supports advanced visualization workflows across the NVIDIA ecosystem, including compatibility with applications built on NVIDIA Omniverse libraries to deliver photorealistic, physically-based rendering and immersive review of simulation results. This allows engineering teams to more easily visualize, communicate, and validate complex fluid behavior as part of broader Digital Twin workflows.
Designed to handle complex, highly dynamic fluid behavior that is difficult to capture with traditional simulation methods, the PAMICS solver utilizes a Lagrangian Smoothed Particle Hydrodynamics (SPH) approach, allowing engineers to simulate fluid dynamics directly from raw CAD geometries without the need for meshing. This accelerates simulation workflows for complex multiphase and free surface flows, particularly in scenarios involving arbitrary motion, fluid-structure interaction and splashing that are difficult or impractical to model with traditional grid-based methods. Critical use cases include:
- Rotating Machinery & Powertrains: Accurately predicting oil lubrication and cooling in complex gearboxes and high-power electric motors without geometry simplification.
- Industrial Mixing & Processing: Modeling multiphase flows, surface tension effects, and non-Newtonian fluids in mixers, agitators, and food processing equipment.
- Water Management: Simulating vehicle wading, soiling, and contamination management for consumer, off-highway and industrial vehicles.
“At SimScale, our mission is to empower engineers to explore thousands of engineering decisions in seconds,” said David Heiny, CEO of SimScale. “By integrating AI Engineering’s sophisticated PAMICS solver, we are bringing a true ‘no-mesh’ workflow to the cloud, removing one of the biggest bottlenecks when it comes to simulating complex fluid dynamics with moving assemblies. Combined with accelerated computing on GPUs and Physics AI workflows, this enables our customers to build their own synthetic data engines, accelerating their path to predictive Digital Twins.
“We developed PAMICS to handle the most demanding fluid dynamics challenges where traditional methods fail—specifically where complex motion and free surfaces interact,” said Dr.-Ing. habil. Stefan Adami, CEO of AI Engineering GmbH. “Joining forces with SimScale allows us to scale this technology globally. As a member of NVIDIA Inception, we have optimized PAMICS to extract maximum performance from NVIDIA GPUs. Delivering this through SimScale’s browser-based platform means engineers in the industrial and manufacturing sectors will be able to access high-end SPH capabilities instantly, without investing in expensive local hardware.”
This integration is being showcased at NVIDIA GTC, attendees can visit SimScale at booth 168 for a preview of the new SPH capabilities running on NVIDIA infrastructure.
