Prospective GSoC 2026 student interested in benchmarking and performance optimization

[MakiBuu]

Hello everyone,

My name is Maren and I’m a computer science student currently studying in Spain. I recently started looking into the Google Summer of Code projects and MDAnalysis immediately caught my attention because of the combination of scientific computing and Python development.

Most of my experience is in Python development, along with Java and C, and I regularly work in Linux environments using Git. I’m particularly interested in performance optimization, data processing, and building tools that help other developers and researchers work more efficiently.

After reading through the proposed GSoC projects, the “Benchmarking and performance optimization” project stood out to me. I find the idea of improving performance visibility across the codebase and identifying bottlenecks very interesting. Writing benchmarks that help guide future optimizations also seems like a very practical contribution to the project.

Over the next days I plan to:

• set up the development environment locally
• explore the existing ASV benchmarks
• get familiar with the parts of the codebase that are currently covered by benchmarks

Before diving in, I wanted to introduce myself and ask if there are particular parts of the library where additional benchmarking would currently be most helpful. If there are beginner-friendly issues or areas where benchmark coverage is missing, I would really appreciate being pointed in the right direction.

Looking forward to learning more about the project and hopefully contributing.

Best regards,
Maren Salanova

[orbeckst]

You have a good plan for first steps. I'd recommend writing a simple benchmark — just choose anything that's not covered already. It's more important to learn the process than to be exhaustive. As a recent example you can have a look at PR #5277

The MDAnalysis.lib module has some code that is used repeatedly. All coordinate readers/writers are important. Many analysis tools are not covered and are good targets.

[MakiBuu]

Hi, thanks for the suggestions!

I’ll start by going through the current ASV benchmarks and then try to add a simple one for something that isn’t covered yet. I’ll also check PR #5277 to get a better idea of how benchmarks are typically structured.

If anything comes up while setting things up or writing the benchmark I’ll ask here.

Thanks again for the guidance!