Often the Revit product team is asked how the Revit product line takes advantage of multiple processors and the workstation GPU, so we thought we'd provide a list for your reference, and a video that shows a new feature of Revit 2016 that uses multiple processors to accelerate model navigation to speeds comparable to the Navisworks 2016 model.
Our thanks to the Beck Group for providing the Sa Rang project model above as an example of a large project to test these capabilities in development.
A number of high-impact areas in Revit use multiple processors to improve performance and model scalability anywhere from 10 – 1500%, depending on the operation.
Multi-core functions in Revit:
- Progressive display – new in 2016 – used by all models all the time
- Vector + Raster Printing
- 2D Vector Export such as DWG and DWF
- Mental Ray Rendering
- Wall Join representation and update in plan and section views
- Loading elements into memory, reducing view open times when elements are displayed for the first time in a session.
- Parallel computation of silhouette edges such as outlines of a curved surfaces, used to accelerate navigation of perspective views.
- Translation of high level graphical representation of model elements and annotations into display lists optimized for graphics cards, engaged when opening views or changing view properties.
- File Open and Save
- Point Cloud Data Display
Revit GPU Use
Revit uses GPU processing for drawing all graphics through OGS and DirectX 11, giving Revit approximately a 30% increased drawing operation speed over Microsoft’s simulated CPU graphics engine, Warp.
In 2015 and 2015 R2 Revit added a number of new enhancements to increase GPU use:
- Consolidation - draws multiple geometries of the same material to the GPU as a batch
- Instantiation - draws multiple identical objects to the GPU as a batch