https://github.com/abhiroop/snape

A distributed scheduler. [WIP]
https://github.com/abhiroop/snape

distributed-computing haskell parallel-computing

Last synced: 11 months ago
JSON representation

A distributed scheduler. [WIP]

• Host: GitHub
• URL: https://github.com/abhiroop/snape
• Owner: Abhiroop
• License: bsd-3-clause
• Created: 2018-01-17T21:15:47.000Z (about 8 years ago)
• Default Branch: master
• Last Pushed: 2018-02-15T00:49:52.000Z (almost 8 years ago)
• Last Synced: 2025-01-24T18:46:21.416Z (about 1 year ago)
• Topics: distributed-computing, haskell, parallel-computing
• Language: Haskell
• Size: 77.1 KB
• Stars: 2
• Watchers: 4
• Forks: 0
• Open Issues: 5
• Metadata Files:
  • Readme: README.md
  • Changelog: ChangeLog.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# snape

Parallelize your workload. Like magic.

![alt text](https://raw.githubusercontent.com/Abhiroop/snape/master/snape.png "Snape")

*As there is little foolish wand-waving

here, many of you will hardly believe this is magic. - Severus Snape*

### TODOS:

- [x] Use Unagi Chan Queue and replace the naive queue

- [ ] Explore all LINQ/SQL operators and use them as a Task

- [x] Make IO part of the monad transformer stack

- [ ] Performance analysis of Writer monad.

- [ ] Must add **WORK STEALING** from peers

- [ ] Allow dynamic growth of task graph post compilation

- [ ] Design a graph IR (more precisely a DAG)to which the frontend compiles down to

- [ ] Investigate SIMD vectorization options in GHC to enhance Repa

- [ ] Add documentation and examples

### INSPIRATION

- Naiad - Timely Data Flow https://www.microsoft.com/en-us/research/publication/naiad-a-timely-dataflow-system-2/

- http://dask.pydata.org/en/latest/