Orbital data center feasibility (spacedc-mdao)
- Python
- OpenMDAO
- pymoo
- SALib
- Streamlit
Problem
“Put the data center in orbit” is a popular idea, and the popular version skips the engineering. Power generation, thermal rejection, network capacity, and reliability all trade against each other, and nameplate specs rarely survive that interaction.
Approach
This package models the problem with OpenMDAO and treats it as a multidisciplinary optimization, not a single spreadsheet. It runs a degradation waterfall from nominal specs down to delivered capacity, with constraints on power, thermal, network, and reliability, plus sensitivity analysis to see which assumptions actually drive the result. A Streamlit app exposes the trade space.
Result
The analysis quantifies how far real delivered capacity sits below the nominal pitch and where the binding constraints are. It is the kind of systems-level, cost-aware modeling that turns a confident claim into a feasibility number you can argue with. Try the live demo or read the code on GitHub.