[challenge]: Port any challenge of this school to Rust

Released by

Hiroshi Shinaoka, Saitama University

Contact email

shinaoka@mail.saitama-u.ac.jp

Method

Other

Challenge issue

Context. Rust has long been seen as a better C++ for scientific computing: memory safe by construction, with modern tooling. What held it back was the steep learning curve, and agentic AI has removed that barrier. What is still missing is the library ecosystem. This challenge probes it, using the other challenges of this school as the test set.

The challenge. Pick any challenge of this school. Solve it with whatever established stack you like (or reuse your group's solution), then port the solution to Rust. Where tensors are involved, tenferro-rs, an experimental pure-Rust tensor library (einsum, linear algebra, autodiff, CPU/GPU), is a natural substrate, and its gaps are exactly what we want to find. The original implementation is your oracle and benchmark reference; you never need to trust the Rust stack, you measure it.

Deliverables, in three tiers, all valuable.

1. A working Rust port reproducing the reference observables within stated tolerances, plus a benchmark table against the original implementation.
2. A gap list: missing operations, numerical discrepancies, or performance gaps found while porting. For tenferro-rs, fix each into the canonical record of the verification challenge (#114): oracle cases and benchmark entries. For the wider Rust ecosystem, a written survey of what is missing is itself a deliverable.
3. A porting report: API friction and design feedback, written for maintainers.

An incomplete port with a high-quality gap list is a success, not a failure.

Scoring and recognition. Same as #114: systematic over single, merged canonical artifacts over reports, contributor listing, and co-authorship qualification on the planned tenferro-rs software paper for substantial and sustained contributions to the tenferro-rs ecosystem.

Why this may lead to research output. Each port is a case study of agentic migration of scientific code to Rust, and the collected gap lists define, empirically, what a scientific Rust ecosystem must provide. The tenferro-rs findings feed directly into the planned software paper.

Start by selecting one of the school's challenges and reading its reference implementation, then review tenferro-rs and the verification challenge in #114. Reproduce the reference observables in Rust, record benchmark results and any library gaps, and provide the requested porting report; the issue does not identify a specific challenge or repository file.