Benchmarks before and after scaling: showcasing formulation and solver tuning improvements

[danielelerede-oet]

Summary

Idea proposed by David Paolella during discussions with @pz-max on how benchmark results could be used not only to compare solvers, but also to educate model developers on formulation and scaling best practices: create a dedicated benchmark website section demonstrating how optimization models can be transformed from slow, memory-intensive, or unsolved instances into tractable problems through improvements in model formulation, scaling, and solver configuration.

To be combined with PyPSA - Roadmap (view)

Deliverables

• Website section
• Blog post (LLM-readable and human-readable)
• Webinar/workshop for model developers
• Recorded presentation with reproducible examples