Supercharge Your Engineering with Advanced AI
Accelerate R&D and Optimize Processes Using Ansys SimAI or Stochos
Push your engineering to new frontiers with Ansys SimAI in the cloud and Stochos on-prem machine learning, each fueling unstoppable innovation, deeper insights, and faster results. Find the ideal solution—or harness both—for your unique challenges. Whether you need a cloud-powered engine to supercharge simulation workflows (Ansys SimAI) or on-prem control for highly sensitive or varied data (Stochos), our experts will guide you every step of the way.
Two Breakthrough AI Solutions for Unmatched Speed & Scale
Ansys SimAI in the cloud and Stochos on-prem machine learning
Ansys SimAI: Cloud-Powered Intelligence to Transform Your Simulations
- Accelerated Model Training: Leverage existing simulation outputs to train AI in days, then predict results in seconds.
- Multi-Physics & Complex Geometries. Handle fluid, structural, thermal, and electromagnetic data with minimal manual prep.
- Expand Your Exploration – Evaluate dozens more design variants without ballooning HPC costs or schedule delays.
Stochos: The On-Prem AI Engine for Real-World Innovation
- End-to-End Machine Learning, In-House: Keep all your data and intellectual property on site while deploying advanced algorithms.
- Generative Design & Bayesian Optimization – Rapidly explore new shapes or parameters and converge on the best solutions faster.
- Efficient & Adaptable – Minimal hardware requirements, seamless integration, and quick learning from small or large datasets.
Ansys SimAI: Cloud AI that Supercharges Simulation
Ansys SimAI uses your existing simulation data to train an ultra-fast, high-fidelity AI model. From fluid and thermal to structural and electromagnetic, it tackles all physics—unlocking more design exploration at a fraction of the time and cost.
- Train in Days, Predict in Seconds – Turn historical solver outputs into quick AI-driven insights.
- No Parametrization Hassles: Handle inconsistent geometries and multi-physics seamlessly.
- Accelerate Design Exploration. Evaluate up to 50× more configurations.
- Reduce HPC Overhead – Run fewer full simulations while achieving top-tier accuracy.
- Cloud-Collaborative: Give distributed teams easy access to AI models and results.
Stochos: On-Prem AI that Redefines Engineering Boundaries
Stochos transforms your engineering data—no matter how diverse—into actionable insights. With minimal hardware demands and powerful generative features, it unlocks rapid concept exploration while keeping your intellectual property safely on-site.
- DIM-GP Algorithm – Handles noise and non-linearities without specialized tuning.
- Generative AI & Bayesian Optimization: Quickly propose new shapes or parameters.
- Ultra-Fast Predictions. Seconds vs. hours of traditional HPC.
- Streamlined Integration – Export as interactive web apps or plug directly into workflows.
- 100% On-Prem: Maintain full control and security over your data.
Ansys SimAI Case Studies
Ansys SimAI In Action
Bumper Impact Performance
Challenge
- Bumper design is driven by several factors: A lightweight design improves range and fuel efficiency.
- Safety regulations for cars require the evaluation for robustness and manufacturing variations. Performing physical crash tests are very expensive and time consuming.
- A Multi-disciplinary virtual optimization approach is needed to get the best performance in safety, durability and NVH.
Solution
- ~ 50 different crash models with varying part thicknesses were evaluated to generate a surrogate AI model for bumper impact. (SimAI, LS-DYNA)
- SimAI model accurately predicts the bumper deformation and barrier forces as a transient response. (SimAI)
- SimAI Prediction on new bumper thickness in less than 1 min. Overall crush predictions have an error of less than 0.5% and barrier force error is within 10%. (SimAI)
Benefits
- Assess more car designs: 20x compared to traditional simulation methods and optimize quicker.
- Predict consistent safety performance across design changes faster (>50x compared to classical crash simulation), even when the geometry structure is inconsistent by leveraging on past crash database (earlier design phase, previous car generations).
- Shift Left: Cut-down your design process duration and cost by allowing designers to use fast and meaningful crash prediction.
Bumper Impact Performance
-
Traditional Crash Analysis: Dozens of hours for each variation
-
SimAI: ~1 min per variant, <0.5% crush error
-
Outcome: 20× more designs tested, ensuring safety standards faster.
SUV Aerodynamic Performance
Challenge
- CO2 emission reduction plans (e.g. WLTP) require automotive manufacturers to assess the aerodynamic performance of all design variants of a new car family.
- Wind tunnel tests are too slow and costly to solve this challenge. And it is important to shrink the development time of the next generation of electric cars.
- Drastically faster predictions of aerodynamic performance are required to stay competitive.
Solution
- ~ 50 accurate CFD results are used to create the AI model, including car exterior shape variations and topological changes (rear mirror, ski rack, spoiler, etc). (Fluent, SimAI)
- SimAI Prediction on new SUV geometry in less than 1 min. (SimAI)
- SimAI Drag error compared to CFD: less than 0.5% (5 to 10 drag counts) and accurate skin friction field and wake topology prediction. (SimAI)
Benefits
- Assess more car designs: 20x compared to traditional simulation methods and optimize quicker.
- Predict consistent aero performance across design changes faster (10 to 100x), even when the geometry structure is inconsistent by leveraging on past CFD simulations database (earlier design phase, previous car generations).
- Shift Left: Cut-down your design process duration and cost by allowing designers to use fast and meaningful aero prediction.
SUV Aerodynamic Performance
-
Classical CFD: 5 hours on 200 cores per design
-
SimAI Prediction: ~30 seconds, <0.5% drag error
-
Outcome: Tested 20× more concepts, cutting HPC usage drastically.
Hull Design Exploration
Challenge
- New environmental regulations require to reduce carbon emissions from international shipping (IMO, 2021). One of the methods is to improve ship resistance by optimizing the hull form, which directly translates into energy efficiency and fuel saving.
- For evaluating several concept designs, fast and efficient simulations are needed.
Solution
- 288 accurate CFD results are used to create the AI model, including hull shape variations and operational conditions (draft, boat speed). (SimAI)
- SimAI Prediction on new hull geometry in less than 1 min. (SimAI)
- SimAI resistance error compared to CFD: less than 4% and perfect wave pattern prediction. (SimAI)
Benefits
- Assess the performance of new hull designs in a few minutes instead of days.
- Make informed decisions in the early stages of a design project.
- Save time for value-added tasks: explore a range of solutions and innovate.
Hull Design Exploration
-
Baseline CFD: ~4 hours on 32 cores per run
- SimAI: ~1 minute per new geometry, <4% error
-
Outcome: Identified drag-efficient hull shapes early, boosting fuel efficiency
Stochos Case Studies
Stochos In Action
3D CFD Car Body Pressure
- Without Stochos: ~80 minutes on 16 CPU cores
- With Stochos: ~1 second, ~1.29% mean relative error
- Outcome: Rapid geometry iteration to refine aerodynamics.
FEM Crash Test
- Traditional Simulation: Several hours per scenario
- Stochos Prediction: Seconds, ~5.73% average error across nodes/time steps
- Outcome: Quick safety optimizations, saving HPC resources for final validations
Generative AI for Structures
- Method: Automatically propose new shapes while balancing weight, displacement, and cost.
- Outcome: Engineers find Pareto-optimal designs faster than with brute-force simulation.
