Computational
Design

building sustainable digital workflows

Last updated: 18-10-2026

Department of Architecture · Aalto University

  3. 03Optimization04

    The bridge phase: file weight, translation between software, and computational performance.

    1. 03.01How to reduce file size and stop software from lagging?
    2. 03.02How to extract data (like areas or lengths) into spreadsheets?
    3. 03.03How to check your model before rendering or exporting drawings?
    4. 03.04How to fix broken meshes and naked edges?
← Back to index
03 - OptimizationpublishedUpdated 21 Sept 2026

How to reduce file size and stop software from lagging?

Edited Screencapture from Rhinoceros 7

As a project grows, files fill up with unnecessary geometry and heavy textures. Keeping them lean prevents crashes and lag, and makes working in a team much easier.

Clean Up Your File

Deleting objects doesn't always shrink the file: unused layers, block definitions and materials stay behind, and hidden objects still count. Purge removes what is unused; SaveSmall saves without display data Rhino can rebuild.

You may notice:

  • File size keeps growing even after deleting objects
  • Saving and opening the file take a long time

Search for

Tutorial

Split Large Models with Worksessions

Instead of one huge file, link several smaller files in a Worksession. Each person edits their own file and sees the others' parts as locked references, so the main file stays light.

You may notice:

  • One giant file that everyone needs to edit
  • Only one person can work on the model at a time

Search for

Tutorial

Reuse Geometry with Blocks

A block stores a repeated object only once; every copy is just a reference to it (an instance). A hundred identical chairs then take up barely more space than one.

You may notice:

  • The file grows with every copied element
  • Updating a repeated element means editing every copy

Search for

Tutorial

Simplify Imported Assets

Downloaded models and textures are often far more detailed than you need. Reduce them first — fewer mesh faces (ReduceMesh), smaller texture images — then bring them into your main file.

You may notice:

  • The file gets heavy right after importing a downloaded model
  • The viewport lags after adding textures

Search for

Tutorial

NURBS vs. Mesh

NURBS are precise and editable but heavy; meshes are made of small flat faces and display faster. Keep NURBS for what you are designing, meshes for reference objects, and low-poly meshes for distant context.

You may notice:

  • The viewport lags when the site context is visible
  • Background objects slow the model down as much as your design

Search for

Tutorial

No video yet