2026 R1: Introducing Ansys GeomAI Software and a Reimagined Ansys SimAI Portfolio

Artificial intelligence (AI) is rapidly reshaping how engineers design, explore, and validate product concepts. With 2026 R1, Ansys, part of Synopsys, is accelerating that transformation. This release introduces the Ansys GeomAI AI platform for geometry, a breakthrough generative design technology that learns from reference geometries and produces new concepts grounded in real engineering constraints. Alongside GeomAI software’s debut, Ansys is launching a restructured Ansys SimAI AI platform for simulation, with the introduction of SimAI Pro software for local, component-level experimentation and SimAI Premium software for cloud-scale predictive modeling, giving teams a clear path from first AI trials to full enterprise adoption. Together, these innovations expand what engineers can imagine, create, and evaluate, all in a unified AI-powered workflow.

• SimAI Pro software is new. Teams can now train models and predict physical phenomena on the surface of product components using a desktop machine. SimAI Pro software runs locally with or without a graphics processing unit (GPU), though a GPU is strongly recommended. When projects outgrow the limitations of a desktop machine, you can move the same datasets and models to SimAI Premium software for 3D fields and cloud-scale throughput without reprocessing. This tiered approach lets organizations explore AI on small projects, prove value, and scale when ready while using a consistent user experience and data pipeline.

What’s New in SimAI Premium SaaS

2026 R1 brings focused enhancements that make model building faster, training more robust, and administration easier.

Model build and post-processing

• A build panel now lets users track models as they are created, follow progress, and quickly compare alternatives.
• Nonparametric optimization gains the ability to enforce axial symmetry, helpful for turbomachinery, rotating parts, and other symmetric components.
• Global coefficients are now compatible with point variables, supporting training and prediction on scalar values. This adds flexibility when you need to mix field data with key scalar metrics.

Training data handling

• Support for training datasets larger than 15TB.
• Improved scalability for high-resolution archives, so you can train on richer meshes and finer time steps without manual down-sampling.
• An enhanced pipeline for large‑volume ingestion and processing, reducing the time between data upload and a ready-to-use model.

User interface (UI) and user management

• An announcement banner inside the web app surfaces important product updates right where teams work.
• An expanded user profile centralizes personal information and preferences.
• New Owner and Admin roles provide clearer governance so organizations can manage access and compliance across projects.

Together, these changes push SimAI Premium SaaS toward a more transparent, scalable, and governable platform for AI-powered engineering at enterprise scope.

Introducing SimAI Pro Software: Desktop AI That Meets You Where You Work

Some teams want to test AI on a single component or a small dataset before they consider cloud deployment. SimAI Pro software provides that on-ramp. Install it on a workstation with a compatible GPU, load your existing simulation results, and start training. Engineers can:

• Explore surface predictions for component-level studies.
• Keep data local while evaluating feasibility, accuracy, and workflow fit.
• Avoid machine learning (ML) complexity with guided model creation.
• Upgrade seamlessly to SimAI Premium software for full 3D fields and scale when the project warrants it, without repeating data prep.

The result is a lower barrier to entry and a shorter path from “Let’s try AI” to “This is a standard step in our workflow.”

GeomAI Software: From Reference Designs to New Concepts

If SimAI software answers “How will this design perform?” then GeomAI software helps ask and explore “What does our next design look like?” It learns from explicitly provided design data, creates a latent representation of those geometries, and then generates new concepts inspired by the design language in the training set.

What does this mean for teams searching for new and improved designs?

• Learn from reference geometries to create novel ideas without setting up complex parametric trees.
• Support topology variation and structural changes that are cumbersome with traditional parameterization.
• Preserve design language so generated concepts stay on brand and remain manufacturable.
• Deploy in multiple ways, from manual concept exploration to closed-loop optimization, where an optimizer searches the latent space for designs that meet performance targets.

How it works

GeomAI software builds a parametric latent space that captures the geometry design space of the training dataset. Users can generate new geometries by specifying values in that space. To assess performance, connect the generated geometry to your analysis method of choice — for example:

• SimAI software for rapid predictive evaluation.
• High‑fidelity Ansys solvers when physics detail is essential.
• A response surface mapping latent parameters to known performance metrics for fast screening.

The practical outcome is a true generative loop: GeomAI software proposes new concepts, Ansys optiSLang process integration and design optimization software drives optimization across the latent space, and your analysis tool evaluates performance, all grounded in real data and engineering constraints.

What To Expect From Ansys AI 2026 R1

2026 R1 represents a major leap in Ansys’ AI capabilities, anchored by the introduction of GeomAI software, which brings true generative design to engineering teams. By learning from existing geometries and proposing new, manufacturable concepts, GeomAI software opens the door to rapid exploration far beyond traditional parametric approaches. At the same time, the refined SimAI Pro and Premium offerings give organizations a scalable way to adopt predictive AI, from desktop pilots to cloud-scale 3D physics predictions. Together, GeomAI software and the new SimAI portfolio form a powerful ecosystem: one that helps engineers not only understand how designs will perform but discover what their next design could be.

Get Hands-On: Register for Our Webinars

If you would like to see the workflows end to end, join our two upcoming sessions:

1. SimAI: From Desktop to Cloud With 2026 R1, register here
   
   
2. GeomAI: Generative Design Grounded in Your Data, register here

Where To Start

• If you are curious about AI and want quick results with your own data, install SimAI Pro software on a GPU-equipped workstation and train a model on a recent component study.
• If you already have an archived simulation at scale, move directly to SimAI Premium software and use the improved data pipeline to train on full 3D fields and large datasets.
• If your team is exploring new form factors or wants to broaden concept space, pilot GeomAI software and connect it to SimAI software or your preferred solver to evaluate candidates.

Every team begins in a different place. 2026 R1 is designed to meet you there and grow with you, with a clear path from first experiment to standard practice.