SimScale
Overview
Company Profile
SimScale is a cloud-based engineering simulation company founded in 2012 in Munich, Germany, by five graduates of the Technical University of Munich (TU Munich): David Heiny, Vincenz Dölle, Johannes Probst, Alex Fischer, and Anatol Dammer.[4][8] The company originated from an initial idea in 2011 to provide simulation consultancy services, which evolved into the development of a full cloud-native platform launched in 2013.[4] Headquartered in Munich, SimScale maintains offices in New York and Boston, and employs a global remote workforce, fostering a diverse team that prioritizes customer-centric engineering practices.[9][10] The organization emphasizes diversity, equity, and inclusion to bring varied perspectives to its operations.[9] SimScale supports work-life balance for employees with families through flexible working hours with no core-time restrictions and strong remote/hybrid work options, enabling parents to manage school/nursery drop-offs, family meals, and emergencies without impacting productivity. The company provides direct contributions/subsidies for nursery or kindergarten childcare for every employee's child (referred to as "mini SimScalers"), with the subsidy included in monthly salary payments.[11] A company blog post features testimonials from multiple parent employees (in Germany, UK, India, France) describing how these policies allow integration of family responsibilities with work, such as shared bedtime routines and support during challenging periods like premature births.[12][13] As of recent reports, nearly 50 employees have children, and the culture explicitly promotes compatibility of family and career, contributing to high employee satisfaction in work-life balance.[14][15] SimScale's mission is to democratize engineering simulation through a cloud-based software-as-a-service (SaaS) model, empowering engineers worldwide to optimize product designs more efficiently and accessibly.[5] As of March 2026, the platform is trusted by over 800,000 users globally, with a focus on enterprise adoption through subscription models, API integrations, and partnerships in CAD/PLM ecosystems, and supports key industries including automotive, aerospace, electronics, manufacturing, and civil engineering.[16][17] As of early 2026, SimScale employs approximately 130-160 people globally, with headquarters in Munich and additional presence in New York and Boston. The company has raised around $60 million in total funding across multiple rounds, including a Series C extension in 2021, and secured $26.3 million in debt financing in 2024 to support R&D and expansion. Revenue estimates place the company in the $20-30 million ARR range, reflecting its growth as a B2B SaaS provider in the engineering simulation market. SimScale positions itself as an AI-native platform, fusing Engineering AI and Physics AI with high-fidelity multiphysics solvers to enable autonomous workflows, surrogate modeling, instant parametric optimization, and exploration of thousands of design decisions in seconds.Platform Fundamentals
SimScale is a full-cloud Computer-Aided Engineering (CAE) simulation software designed to enable engineers to perform complex analyses directly through a web browser, thereby eliminating the need for high-end local hardware or extensive software installations.[1] This browser-based accessibility allows users to run simulations on scalable cloud resources, leveraging parallel computing capabilities to handle demanding workloads without the constraints of on-premise infrastructure. The platform supports seamless CAD import from various tools, including SolidWorks, AutoCAD, and Onshape, facilitating quick model preparation and integration into the simulation workflow.[1] At its core, SimScale integrates open-source solvers such as OpenFOAM for computational fluid dynamics (CFD) with proprietary enhancements to ensure reliability, accuracy, and ease of use in a cloud environment. These enhancements include optimized algorithms and user interfaces that streamline solver setup and execution, making advanced simulations more approachable for non-experts. Launched in 2013 as the world's first production-ready Software as a Service (SaaS) application for engineering simulation, the platform emphasizes browser-based 3D visualization and post-processing tools, allowing real-time interaction with results without additional downloads.[9][1] The fundamental benefits of SimScale's architecture include significantly reduced setup time, as users can access the platform instantly without installations, and enhanced collaboration through shared projects that support team-based editing and review. For small and medium-sized enterprises (SMEs), it offers substantial cost savings compared to traditional on-premise software, by providing scalable resources on a pay-as-you-go model and democratizing access to high-fidelity simulations previously limited to large organizations with dedicated IT infrastructure.[1]History
Founding and Early Years
SimScale originated in 2011 as a simulation consultancy founded by five graduates of the Technical University of Munich (TU Munich), who recognized the need for accessible computational engineering services tailored to startups and small teams lacking in-house expertise.[9] The founders—David Heiny, Vincenz Dölle, Johannes Probst, Alexander Fischer, and Anatol Dammer—had backgrounds in mechanical engineering, computer science, and mathematics, with several holding degrees from TU Munich and additional studies at institutions like Georgia Tech.[4] This initial venture emerged as a spin-out from TU Munich, focusing on providing specialized simulation consulting to overcome the high costs and complexity of traditional computer-aided engineering (CAE) tools.[18] In 2012, the company was officially incorporated in Munich, Germany, marking a strategic pivot from consultancy services to developing a cloud-based simulation platform. This shift was motivated by the emerging potential of cloud computing and software-as-a-service (SaaS) models in 2011, which promised to eliminate the hardware barriers inherent in conventional CAE workflows, such as expensive local installations and maintenance.[9] By leveraging cloud infrastructure, the founders aimed to democratize access to high-performance simulations for engineers worldwide, without requiring significant upfront investments in computing resources.[4] A prototype beta version was developed in 2012, initially targeting fluid dynamics simulations powered by the open-source OpenFOAM solver, allowing early users to test web-based CAE capabilities.[18] The platform faced initial challenges in constructing scalable cloud infrastructure to handle compute-intensive tasks and validating solver accuracy in a distributed environment, ensuring results matched those of traditional desktop software without local setups.[19] These efforts culminated in the full public launch in December 2013, positioning SimScale as the world's first fully web-based 3D simulation platform.[20] During this formative period, the team expanded modestly from the five founders to a small core of engineers based in Munich, concentrating on technical development and platform refinement to support broader adoption.[4]Expansion and Milestones
Following its initial beta phase, SimScale publicly launched its free Community Plan on December 2, 2015, offering users up to 3,000 core hours of annual computing power and 500 GB of storage to democratize access to engineering simulations via a web browser.[21][22] This release was supported by a Series A funding round led by Union Square Ventures, which provided capital to scale the platform's infrastructure and user base.[23] In 2017, SimScale raised Series B funding from June Fund to fuel product development and market expansion.[24] During this period through 2020, the company broadened its simulation capabilities, adding advanced thermal analysis features such as thermomechanical simulations for transient heating and thermal shock in early 2016, enabling engineers to model heat transfer in solids and fluids more comprehensively.[25] In January 2020, SimScale secured a €27 million Series C round led by Insight Partners, which accelerated global operations and platform enhancements, including deeper integrations for multiphysics workflows.[26] The Series C funding continued with a €25 million extension in October 2021, co-led by Insight Partners and Draper Esprit (now Molten Ventures), with participation from Earlybird, June Fund, Vsquared Ventures, and Union Square Ventures, bringing the total round to €52 million and supporting further innovation in cloud-native tools.[27] By 2023, SimScale integrated AI surrogate modeling using Graph Neural Networks to accelerate simulations by approximating complex physics on structured data like meshes, reducing computation times from hours to seconds for iterative design tasks.[28] In 2024, SimScale enhanced its platform for real-time automotive design workflows, incorporating AI-powered features for rapid aerodynamics and structural analysis directly within CAD environments, alongside reaching a milestone of over 600,000 registered users worldwide who had completed more than 4 million simulation jobs.[29][30] In December 2024, SimScale signed a €25 million financing agreement with the European Investment Bank to scale its software development, broaden functionalities, and foster deep-tech advancements in the EU.[31] To support its growing North American presence, SimScale expanded its U.S. operations with teams in New York and Boston starting around 2018, enhancing local support and sales.[32] As of 2025, SimScale continued its product evolution with the Summer 2025 update, introducing enhancements like probe points for electromagnetics post-processing and advanced multiphysics couplings, building on prior releases such as updates to temperature-dependent material properties (including solids) in August 2024 and non-Newtonian fluid modeling added in 2023.[33][34][35] In March 2025, SimScale launched the world's first foundation AI model for engineering simulation, focusing on turbomachinery like centrifugal pumps, in collaboration with NVIDIA, enabling near-instantaneous predictions.[36] Strategic partnerships with CAD vendors, including PTC's Onshape for seamless simulation integration and Hexagon for advanced structural analysis, further streamlined workflows for users importing geometry directly into the cloud platform.[37][38] In 2026, SimScale's user base exceeded 800,000 globally. On March 24, 2026, the company released its "State of Engineering AI 2026" report, a global study revealing accelerating AI adoption in engineering design and simulation workflows, with leading organizations embedding AI into core operations to evaluate significantly more design possibilities amid rising product complexity. On March 16, 2026, SimScale announced a strategic partnership with AI Engineering GmbH to integrate the PAMICS solver (Smoothed Particle Hydrodynamics, SPH) into the platform. Boosted by NVIDIA AI infrastructure, this meshless approach eliminates meshing bottlenecks, delivering simulation speeds 10-20x faster than traditional grid-based methods for complex industrial fluid dynamics applications with moving assemblies. This enhances SimScale's capabilities in no-mesh workflows for high-fidelity CFD.Recent developments
In March 2026, SimScale announced a strategic partnership with AI Engineering GmbH to integrate the PAMICS® solver into its platform. Powered by NVIDIA AI infrastructure, this collaboration eliminates meshing bottlenecks and achieves simulation speeds 10-20x faster for complex fluid dynamics applications, enabling high-fidelity, meshless CFD and supporting synthetic data generation for Physics AI and digital twins.[39] On March 24, 2026, SimScale released its 2026 State of Engineering AI Report, a global study surveying engineering leaders. Key findings include that organizations integrating AI with modern cloud infrastructure evaluate significantly more design possibilities (nearly 4x more than non-AI teams), though ambition often outpaces execution due to data silos, legacy tools, and leadership gaps. The report positions SimScale as a leader in bridging these challenges through its AI-native platform.[40]Technology and Features
Simulation Modules
SimScale provides a suite of simulation modules that enable engineers to perform advanced analyses directly in the cloud, leveraging open-source solvers for high-fidelity results without local hardware constraints. These modules cover core engineering disciplines, including fluid dynamics, structural mechanics, thermal management, and electromagnetics, each tailored to specific physical phenomena while supporting scalable parallel computing for complex models.[41]Computational Fluid Dynamics (CFD)
The Computational Fluid Dynamics (CFD) module in SimScale simulates fluid flow, heat transfer, and multiphase flows using the finite volume method, primarily powered by the OpenFOAM solver. This approach discretizes the governing equations over a computational mesh to predict velocity, pressure, and temperature fields in applications such as aerodynamics, HVAC systems, and chemical processes. For incompressible flows, where density variations are negligible, the module handles laminar and turbulent regimes with turbulence models like k-ω SST. Compressible flows, relevant for high-speed scenarios, account for density changes exceeding 30% of the speed of sound. Multiphase simulations employ the Volume of Fluid (VoF) method to model interfaces between immiscible fluids, such as air and water in droplet dynamics. Heat transfer capabilities include convective cooling through natural or forced convection, with radiation modeled via surface-to-surface or discrete ordinates methods. The core governing equations for fluid motion are the Navier-Stokes equations, solved iteratively in the cloud to manage large-scale transient simulations efficiently:
Here, is the velocity vector, is pressure, is density, is kinematic viscosity, and represents body forces; cloud-based solving allows for rapid convergence on distributed resources.[42]
Finite Element Analysis (FEA)
Finite Element Analysis (FEA) in SimScale focuses on structural mechanics, encompassing linear and nonlinear static, dynamic, and thermo-mechanical analyses using the code_aster, CalculiX, and Hexagon's Marc solvers (integrated as of July 2025). These open-source and advanced tools apply the finite element method to discretize solid domains into elements, solving for displacements, stresses, and strains under applied loads and constraints. Static analyses evaluate steady-state responses to forces, pressures, or thermal expansions, supporting both linear elastic materials and nonlinear behaviors like plasticity or large deformations. Dynamic simulations capture time-dependent effects, such as vibrations or impact, through modal or transient solvers that compute natural frequencies and mode shapes. Thermo-mechanical analyses couple structural deformation with temperature-induced expansions, ideal for components under combined thermal and mechanical loading. CalculiX excels in straightforward linear static problems with robust contact handling, while code_aster offers advanced nonlinear capabilities, including cyclic symmetry and reduced integration for efficiency. Marc provides robust handling of large deformations, contact interactions, and material nonlinearities. Material models incorporate temperature-dependent properties, such as Young's modulus varying with heat, to reflect real-world conditions.[43][44][45]Thermal Analysis
The Thermal Analysis module simulates heat transfer processes, including conjugate heat transfer, convective cooling, and radiation, with support for temperature-dependent material properties across solid and fluid domains. Conjugate heat transfer (CHT) models coupled conduction in solids and convection in fluids at interfaces, using either body-fitted meshes or the immersed boundary method (IBM) for complex geometries like heat exchangers or electronics cooling. Convective cooling analyzes buoyancy-driven or forced flows influenced by temperature gradients, incorporating radiation via view factors or Monte Carlo ray tracing for accurate surface heat exchange. Standalone heat transfer in solids computes temperature distributions and fluxes under boundary conditions like fixed temperatures or heat fluxes, accommodating nonlinear materials where thermal conductivity decreases with rising temperatures. These simulations leverage the energy equation alongside momentum solvers, enabling predictions of hotspots and thermal gradients in multiphysics scenarios. Radiation is particularly emphasized in CHT setups to quantify emissive losses in high-temperature environments. As of August 2024, enhancements include temperature-dependent solid properties such as thermal conductivity and specific heat for CHT analyses.[46][47][48][34]Electromagnetic Simulations
Electromagnetic Simulations in SimScale model low-frequency and emerging high-frequency phenomena using open-source codes, focusing on magnetic fields, induction, and electrostatics in devices like electric motors and antennas. Low-frequency analyses, such as magnetostatics for DC currents or time-harmonic magnetics for AC induction, compute magnetic flux density, current density, and inductances in nonlinear materials like ferromagnetic cores. Time-transient magnetics handles varying fields from pulsed currents, supporting applications in switched reluctance motors where core losses and thermal coupling are critical. Electrostatic simulations evaluate electric field strengths and capacitance for insulator designs. High-frequency capabilities remain under development as of September 2025 and will extend to antennas via frequency-domain solvers for wave propagation. These modules integrate with thermal analyses for joule heating effects, using finite element discretization to solve Maxwell's equations efficiently in the cloud. Examples include optimizing motor efficiency by minimizing eddy current losses. Recent additions as of August 2024 include core loss models for electrical steel and power ferrite, nonlinear material support with B-H curves, and probe points for targeted post-processing (Summer 2025).[49][50][34][33]Advanced Tools and Integrations
SimScale incorporates AI surrogate modeling to accelerate engineering simulations by training lightweight models on high-fidelity data, enabling real-time predictions for design optimization such as in turbomachinery applications. These models, developed in collaboration with NVIDIA's PhysicsNeMo framework, can reduce computation times dramatically—for instance, the foundation AI model for centrifugal pump simulations, released in March 2025, speeds up predictions by up to 2,700 times compared to traditional methods—while maintaining causal physics-based accuracy. By leveraging graph neural networks in the backend, SimScale's surrogates approximate complex outcomes like flow routing or structural responses, allowing engineers to explore thousands of design variants in seconds without full recomputation. As of Summer 2025, Physics AI has expanded to the multi-purpose solver for instant outcome predictions and streamlined setup via foundation models.[51][52][36][33] Workflow tools in SimScale enhance simulation efficiency through parametric studies, which enable rapid evaluation of multiple operating conditions or geometric variations to optimize equipment performance.[53] Design optimization is supported via integrated automation, such as linking simulations with external tools like ESTECO for iterative refinement of structural or fluid designs.[54] Automated reporting features streamline result documentation, while specialized controls like probe points facilitate targeted measurements in electromagnetic analyses, such as field intensities at specific locations.[55] Additionally, pocket face selection in CAD preparation automates the identification and grouping of surfaces for boundary conditions, simplifying setup for intricate models.[56] SimScale offers direct API integrations with CAD platforms including Onshape for seamless model import and simulation without file downloads, Autodesk Fusion 360 via a dedicated add-in for one-click geometry uploads, and Siemens NX through compatible neutral formats and API access for workflow automation. As of Summer 2025, integration with nTop supports implicit geometry for CFD applications.[57][58][59][33] These connections support cloud-based collaboration, featuring shared project libraries for team access and version control to track design iterations across distributed users.[60] Recent enhancements from 2024 to 2025 include the integration of Hexagon's Marc nonlinear structural solver in July 2025, which enables robust analysis of large-deformation mechanics, contact interactions, and material nonlinearities in cloud environments; humidity source modeling added to CFD workflows in Q3 2023 for moisture injection or evaporation in multiphysics scenarios like HVAC systems; and robust meshing improvements for complex geometries, with features like automatic CAD surface merging (ongoing). Additional updates encompass porous media modeling and multicomponent gas mixing in CFD (August 2024), viscous heating and turbulence modeling for CHT in thermal analyses (Summer 2025), and pin connectors for structural simulations (August 2024).[45][35][61][34][33] Post-processing in SimScale provides an integrated 3D visualization environment for interactive exploration of results, including isosurfaces, streamlines, and vector fields directly in the browser.[62] Results can be exported in formats compatible with ParaView for advanced offline analysis, supporting detailed filtering and scripting for large datasets.[63]Automated Design Optimization and Workflows
SimScale supports automated design optimization loops primarily through its comprehensive API, which enables programmatic control over simulation setup, execution, and result extraction. This allows integration into closed-loop workflows with external parametric CAD and optimization tools, leveraging cloud parallelism for high-throughput evaluations.API and Integration Capabilities
The API provides access to geometry import, meshing, physics setup, parallel simulation runs, and post-processing outputs (e.g., pressure drop, efficiency, temperature). It facilitates seamless connections with CAD platforms like Onshape and optimization software, enabling fully automated parameter variation, simulation triggering, and feedback for iterative refinement. Key integrations include:- ESTECO modeFRONTIER + VOLTA: In centrifugal pump optimization, Onshape parameterizes geometry (7 variables); modeFRONTIER's MDO reduces ~1 million permutations to 65 using self-adaptive algorithms; SimScale runs parallel incompressible CFD via API, outputting head, efficiency, etc., for feedback loops.[54]
- CAESES (Friendship Systems): For shape optimization, e.g., KSB heat circulator pump impeller (14 parameters). Python scripts push variants to SimScale for DoE CFD (377 variants, >900 simulations across flow rates in ~42 hours parallelized, ~$300 cost), building surrogate models for EEI minimization.[64]
- pSeven Enterprise: Surrogate-based optimization, e.g., IGBT cooling plate; Onshape parameterization, pSeven orchestrates API calls to SimScale for CFD, iterating toward Pareto optima for temperature and pressure drop.[65]
AI-Accelerated Optimization
SimScale's Physics AI deploys surrogate models (pre-trained or user-trained on CAE data) for instant predictions, e.g., evaluating 60+ centrifugal pump designs in under 60 seconds (2700x faster than full CFD). Engineering AI (agentic) orchestrates autonomous workflows via natural language, automating setup, iteration, and documentation for design exploration.[66] These features enable rapid parametric studies and broad design space exploration without native built-in optimizers—logic resides externally, but SimScale provides scalable simulation backend.Data-Driven Model Calibration and Validation
SimScale emphasizes verification and validation (V&V) to ensure simulation credibility, following standards like ASME V&V-20. The platform maintains a public library of validation cases where simulations are compared directly to experimental or analytical benchmarks across various analysis types, including incompressible/compressible flows, multiphase, conjugate heat transfer, structural benchmarks (e.g., NAFEMS), and electromagnetics (e.g., TEAM workshops). These cases calibrate the solvers by quantifying discrepancies and improving accuracy. For data-driven approaches, SimScale's Physics AI enables training of surrogate models from high-fidelity simulation data (minimum 20 runs varying parameters like geometry or boundary conditions). Training uses cloud GPUs; models predict fields (e.g., pressure, velocity) rapidly. Accuracy is assessed via metrics: Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), R² (near 1 ideal), with loss curves to avoid overfitting. Validation compares AI predictions to full solver results; extrapolation warnings enhance confidence. Real-time digital twins integrate Physics AI surrogates with live sensor data via APIs for closed-loop validation, predictive maintenance, and model refinement using operational data. However, built-in tools for automated inverse calibration of underlying physics parameters (e.g., tuning turbulence constants to match experiments) are limited; users typically perform this manually or externally. AI training focuses on simulation-generated data for surrogates, with real-world data integrated indirectly via digital twins rather than for core model calibration.Digital Twins and Virtual Prototyping
Digital Twins
SimScale supports real-time digital twins by combining high-fidelity cloud simulations with AI, enabling Physics AI surrogates—lightweight AI models trained on simulation data—to connect with live operational data for predictive insights, operational reliability, and risk reduction. This creates closed-loop systems where real-world conditions update the digital twin continuously. Key applications include:- Vale: Developed digital twins of large magnetic separators in mining operations to assess structural reliability under harsh conditions, enhancing equipment durability and minimizing maintenance.
- Leaftech: Utilizes SimScale's Lattice-Boltzmann solver (Pacefish®) for microclimate digital twins in built environment design and retrofit, enabling rapid wind flow and thermal simulations on complex geometries with minimal CAD preparation.
Virtual Prototyping
SimScale excels in virtual prototyping by allowing engineers to upload CAD models and run parallel multiphysics simulations (CFD, FEA, thermal) in the cloud, supporting parametric sweeps for thousands of design variants without local hardware limits. This replaces or reduces expensive physical prototypes, accelerating iteration and optimization. Examples of benefits:- Johnson Screens (Aqseptence Group): Verified airflow through an architectural radiator grille via CFD in 18 minutes, avoiding $7,000–$15,000 and months of physical testing.
- Hazleton Pumps: Generated full pump performance curves in minutes using rotating machinery tools—a ~100x speedup compared to traditional methods.
- L&T Construction: Optimized sump pump house structures to eliminate vortices, saving ~15 days and $38,000 in project time and costs.
- General: Users report reductions in physical prototyping costs by up to 45% in CFD/FEA projects, with AI enhancements enabling faster design convergence.
