https://github.com/j-fu/julia-tests
Performance and other tests for Julia
https://github.com/j-fu/julia-tests
Last synced: 2 months ago
JSON representation
Performance and other tests for Julia
- Host: GitHub
- URL: https://github.com/j-fu/julia-tests
- Owner: j-fu
- License: mit
- Created: 2020-02-06T16:33:15.000Z (about 5 years ago)
- Default Branch: master
- Last Pushed: 2021-03-24T13:29:04.000Z (about 4 years ago)
- Last Synced: 2025-01-08T12:39:44.693Z (4 months ago)
- Language: Julia
- Size: 399 KB
- Stars: 0
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Julia tests
## Parallelization test
See subdirectory `parallel` .
### Rationale
Iterative methods for PDEs use vector operations: basic vector algebra, scalar products, sparse matrix - vector multiplication.
For large problems, these are characterized by a high ratio of memory access vs floating point operation. Here we use the
"Schönauer Vector Triad" in order to compare performance for various implementations, and compare Julia to C:
````
for i=1:N
d[i]=a[i]+b[i]*c[i]
end
````
The use of this benchmark has been inspired by the [benchmarking site of Georg Hager](https://blogs.fau.de/hager/archives/tag/benchmarking)
This benchmark of course reveals the various memory access issues of modern processors. In addition it can reveal implementation issues for parallel methods.