ODYSSEE Solver 2026.2 Expands Predictive Modeling and AI Performance

Engineering organizations are generating more data than ever before. Simulation results, test measurements, image data, and operational information all contain valuable insights that can help improve product performance and accelerate innovation. The challenge is no longer collecting data. The challenge is transforming that data into actionable engineering knowledge.

This is where ODYSSEE Solver comes in.

ODYSSEE Solver is the AI and analytics engine of the ODYSSEE portfolio. It enables engineers and data scientists to transform simulation, test, and operational data into predictive models, optimization strategies, and actionable insights. Combining machine learning, data mining, predictive modeling, scientific analysis, and workflow automation in a single environment, ODYSSEE Solver helps organizations accelerate engineering decision-making and extract greater value from existing data assets. It includes the Quasar AI engine for predictive analytics and data mining, as well as the Nova optimizer for design exploration and optimization.

Whether the goal is understanding complex engineering behavior, identifying trends within large datasets, building surrogate models, or finding optimal design parameters, ODYSSEE Solver helps bridge the gap between raw data and informed decision-making.

What’s New in ODYSSEE Solver 2026.2?

With the release of ODYSSEE Solver 2026.2, Cadence introduces several enhancements that further strengthen its predictive modeling and AI capabilities.

Expand Your Predictive Modeling Toolbox

One of the most significant additions in ODYSSEE Solver 2026.2 is the introduction of Decision Tree Regressor and Random Forest Regressor capabilities. These new tree-based machine learning methods provide additional options for building predictive models from engineering data.

Decision Tree Regressor partitions the data space into segments based on variance minimization, making it easier to capture complex relationships between inputs and outputs. Random Forest Regressor takes this approach further by combining multiple decision trees trained on slightly different versions of the dataset, resulting in more robust and accurate predictions.

For engineers working with simulation results, test data, or manufacturing information, these methods provide greater flexibility when selecting the most suitable predictive modeling approach.

New Regression Methods for High-Dimensional Data

Engineering datasets often contain hundreds or even thousands of variables. Understanding which parameters truly drive performance can be difficult.

To address this challenge, ODYSSEE Solver 2026.2 introduces three powerful regularized regression methods:

• LASSO Regression
• Ridge Regression
• Elastic Net Regression

These techniques are particularly useful for datasets with many variables because they help simplify models while preserving predictive accuracy. By automatically focusing on dominant variables, engineers can create more interpretable models and better understand the factors influencing engineering performance.

Accelerate AI Workflows with GPU Computing

As machine learning and deep learning projects become more sophisticated, computational performance becomes increasingly important.

ODYSSEE Solver 2026.2 now offers a dedicated CUDA-enabled version for Windows, allowing supported workflows to leverage GPU acceleration.

GPU acceleration is available for several machine learning and deep learning methods, including MLP, Modified U-Net, GLTSM, LSTM, and transfer learning workflows. By utilizing GPU hardware, users can significantly accelerate training and computation times, enabling faster experimentation and more efficient development of AI-driven applications.

For organizations exploring advanced predictive analytics, image-based learning, and neural network workflows, this enhancement can help reduce turnaround times and improve overall productivity.

Greater Flexibility for Result Evaluation

ODYSSEE Solver 2026.2 also introduces the new Compare1 function. This function calculates the same evaluation criteria as the existing Compare function while returning raw, untransformed results.

This gives engineers and analysts greater flexibility when creating custom reporting, validation, and post-processing workflows. By providing direct access to raw comparison results, Compare1 allows users to further tailor analyses to specific project requirements.

Enabling Faster Engineering Decisions

The latest enhancements continue ODYSSEE Solver's mission of helping organizations transform engineering data into actionable insights. The addition of tree-based predictive models expands the available machine learning toolbox. New regularized regression methods improve the analysis of complex datasets. GPU acceleration enhances AI performance. And the Compare1 function provides greater flexibility for result evaluation.

Together, these capabilities help engineers build more accurate predictive models, explore more design alternatives, and accelerate engineering decision-making across simulation, testing, manufacturing, and product development workflows.

Ready to unlock more value from your engineering data? Visit the ODYSSEE product page to learn more about ODYSSEE and its AI-powered engineering analytics capabilities.