SymPy Performance Proposal for GSOC – Short Summary & Wiki Link

[fırat]

Dear Maintainers,

I’d like to share my performance-focused GSoC proposal for SymPy.

Originally, I planned a SymEngine integration proposal (which was removed from the ideas list). I then discussed this idea with Ondřej Čertík, who indicated that he could mentor such a project. However, I understand that other maintainers have reservations about SymEngine. Therefore, I wanted to expand my proposal to include broader Python-based optimizations and optional backends (e.g. Python-Flint) so that we can tackle performance more holistically.

A few days ago, I asked for permission to create a Wiki page, and Aaron kindly approved it.

You can find my proposal on the SymPy Wiki here: [Wiki Proposal Link].
I would greatly appreciate any feedback on which areas (e.g., polynomial routines, domain usage, matrix ops, etc.) should be prioritized, or how best to incorporate optional backends without complicating SymPy’s core. 

I recognize that the performance improvements outlined in this proposal touch on multiple parts of SymPy. Although the project is potentially large, I plan to break it down into manageable milestones for GSoC, focusing on the most impactful and realistic goals first. I’d love to have your input on which sub-tasks are highest priority or most beneficial to the community, so I can ensure the final deliverables remain both achievable and valuable within the GSoC timeframe.

Thank you in advance for your time and insights. I look forward to refining this project based on your suggestions!

Best regards,
Fırat Bezir
(frt...@gmail.com)