...
 
Commits (1)
...@@ -150,6 +150,30 @@ If you add a benchmark try to set the problem sizes for ...@@ -150,6 +150,30 @@ If you add a benchmark try to set the problem sizes for
fast/normal/slow reasonably. [Modes](#modes) lists the recommended brackets for fast/normal/slow reasonably. [Modes](#modes) lists the recommended brackets for
each mode. each mode.
### Categories
When you add a benchmark, you'll have to make a conscious decision in which to
(historical) category to put it. Here's a description for each, taken from the
original [Nofib paper](https://link.springer.com/chapter/10.1007%2F978-1-4471-3215-8_17):
#### Single-threaded
This is the default mode you want to add to most of the time.
- `imaginary`: A toy benchmark. Solvers for puzzles like n queens.
- `spectral`: Algorithmic kernels. If you have a library out of which you have
extracted a core algorithm, put it here.
- `real`: Actual applications from the "real world". Because most applicationsof today depend on quite many packages, everything in here is rather dated.
- `shootout`: The (toy) benchmarks from the [benchmarks game](https://benchmarksgame-team.pages.debian.net/benchmarksgame/), formerly known as the language shootout.
#### Other modes
In addition to the above single-threaded mode, you can also add your benchmarks to one of the
I (SG) haven't really run these, yet, but
- `gc`: Measures the GC.
-
### Stability wrt. GC paramerisations ### Stability wrt. GC paramerisations
Additionally, pay attention that your benchmarks are stable wrt. different Additionally, pay attention that your benchmarks are stable wrt. different
......