OffchainLabs / OffchainLabs/hashtree
investigate AMD Shani loops
Open
Nobody has claimed this yet.
- Dominant language
- Assembly
- Stars
- 42
- Forks
- 14
- Avg merge
- 11d 15h
- Merged PRs (30d)
- 2
Description
The following is a bench on intel
[ RUN ] shani.shani
[ OK ] shani.shani (mean 2.513ms, confidence interval +- 0.416409%)
[ RUN ] shani.shani_one_at_time
[ OK ] shani.shani_one_at_time (mean 2.588ms, confidence interval +- 0.422037%)
And these are on AMD:
[ RUN ] shani.shani
[ OK ] shani.shani (mean 2.271ms, confidence interval +- 0.119590%)
[ RUN ] shani.shani_one_at_time
[ OK ] shani.shani_one_at_time (mean 3.947ms, confidence interval +- 0.063440%)
They do not happen always, but there is a big difference between the one_at_time bench and the normal run. This was only reproduced on AMD
Contributor guide
No contributing guide indexed for this repository
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 locating and running the shani.shani and shani.shani_one_at_time benchmarks, then reproduce the Intel and AMD results described in the issue. Compare the two benchmark paths and determine the cause of the AMD-only timing difference; done means the discrepancy is explained and any needed follow-up is clearly identified.
Written by the indexing model from the issue text.
Assessment
- Domain
- performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100