Consider physical swapping in AlignedBitMatrix
Nobody has claimed this yet.
- Dominant language
- Rust
- Stars
- 64
- Forks
- 16
- Avg merge
- 1d 5h
- Merged PRs (30d)
- 5
Description
Currently, AssignedBitMatrix uses a rows vector to keep track of row pointers. This means that row swaps are cheap, but also means that other operations cannot assume a contiguous memory layout (see https://github.com/microsoft/qdk-ec/pull/44). It would be useful to investigate the tradeoffs of doing physical swaps. If the swap involves two rows already in cache (as may be the case during echelonize), then it will remain cheap, and could dramatically improve performance of other operations.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by inspecting AssignedBitMatrix's rows vector and the row-swap behavior discussed in pull request #44. Compare physical row swapping with pointer swapping during echelonize and other operations, then document the performance tradeoffs and a justified implementation direction.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- performance
- Issue type
- Refactor
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100