Here's what's new in our latest product update:
We’ve released a Blender integration that enables you to import and parametrically modify Blender models directly in the Luminary platform. With this capability, you can easily generate geometry variations that can be simulated at scale to generate datasets for physics AI model training.
Explore the Blender integration tutorial here.
Configurations allow you to easily swap between different, previously set-up geometry variations. When you create a Configuration, you set up each configuration as a set of 0, 1, or more volumes and give it a unique name.
You then choose an active configuration. All volumes associated with the active configuration are kept, while volumes not associated with the active configuration are removed from the model. You can later choose a different active configuration, and the model will be updated to display the volumes associated with that configuration instead.
We've added a button to the Import panel that allows you to swap the file to easily swap in a different geometry file.
Here's what's new in our latest product update:
Seamlessly import your Blender models into Luminary with full parametric control. This powerful integration preserves shape keys across multiple file formats—allowing you to modify geometry directly within the Luminary platform. No more switching between applications to test design variations.
With Blender's flexible modeling capabilities, you can now rapidly explore design alternatives, run optimizations, and generate rich datasets for physics AI model training. Simply import your parameterized model, adjust values using intuitive sliders, and instantly see how changes affect performance.
Quickly iterate on your designs with our new file swapping functionality. You can now swap imported CAD files with a single click using the new swap button, eliminating the need to start over when you want to try a different model. This approach saves you time and keeps your workflow uninterrupted as you explore design alternatives.
The swap feature works seamlessly with both continuous and discrete geometry types, giving you flexibility with various CAD formats. Currently, swapping from continuous to discrete geometry isn't supported, but we're working on enabling automatic transitions between these types in an upcoming update.
Here's what's new in our latest update:
You can now generate images directly from the Luminary SDK, enabling you to automate the post-processing visualization pipeline in Luminary.
Read more about the latest SDK release here. Note: this release contains breaking changes, please upgrade to the latest version and view the migration guide to update your scripts.
Here's what's new in our latest update:
On the Results page, you can now customize which columns are displayed and rearrange their order. We've also introduced support for column groupings.
Here's what's new in our latest release:
You can now generate Luminary SDK code using the AI Assistant. The Assistant has access to your project metadata, so you can ask it to generate SDK code to perform a variety of tasks without needing to manually enter the parameters. Read more about Luminary's SDK here.
Here's what's new in our latest update:
We're excited to introduce enhancements to the Results page that make working with your simulation results faster and more intuitive:
Navigation & Customization
Contextual Actions
DoE Visibility and Management
We've improved 2D XY plots to provide better visual feedback when analyzing data:
Here's what's new in our latest release:
The latest SDK release includes a number of enhancements to improve the experience of working with our SDK. This release includes breaking changes, so please refer to the migration guide to update your scripts accordingly. Read more about this release here.
You can now include the residuals in the volume solution to assess local convergence and identify problematic regions in the domain.
Here's what's new in our latest update:
Here's what's new in our latest update:
You can now create an API key to authenticate with the SDK without needing to interactively log in. Read more about authentication with API keys here.
Here's what's new in our latest release:
The new Workflow UI is now the default for all users. Based on user feedback, we have sunset the old UI while consolidating its core benefits into the new experience for a more streamlined workflow.
You can now initiate an adjoint analysis directly from the Advanced Analysis dropdown (accessible via the gear icon next to the Run Simulation button).
The LMA control volume (CV) count scheduling has been enhanced with a new "pitch-pause" stair-step approach. This approach optimizes the target CV count for each adaptation and implements repetitions. This leads to better convergence in key quantities of interest and higher overall accuracy, even with a reduced target CV count.
As a result, we now recommend reducing the target CV count by 50% (e.g., if previously 20M, now 10M), which will result in improved accuracy while preserving the same time to solution and cost as the previous version of LMA. This will also lower storage requirements due to the decreased CV count.
Previous Target CV Scheduling:
New "Pitch-pause" CV Target Scheduling:
The latest SDK release includes a number of syntax simplifications and streamlined helper functions to improve the experience of working with our SDK. This release includes breaking changes, so please refer to the migration guide to update your scripts accordingly. Read more about this release here.