-
Notifications
You must be signed in to change notification settings - Fork 59
docs: add benchmark results matrix to README #769
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Changes from all commits
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -22,6 +22,25 @@ uv run examples/ivp_homogeneous_medium.py | |
|
|
||
| No GPU required — all examples run on CPU with NumPy. | ||
|
|
||
| ## Benchmarks | ||
|
|
||
| Reference runtimes for the 3D scaling benchmark in [`benchmarks/benchmark.py`](../benchmarks/README.md). Values are total elapsed seconds for a single default run (3D initial-value problem, heterogeneous absorbing medium, 1000 timesteps, averaged over 3 repeats). | ||
|
|
||
| | Backend | OS | Accelerator | 64³ | 128³ | 256³ | Hardware | | ||
| |---------|-----------------------|-------------------|-----|------|------|----------| | ||
| | python | Linux | CPU | — | — | — | — | | ||
| | python | Linux | NVIDIA GPU | — | — | — | — | | ||
| | python | macOS (Apple Silicon) | CPU | 81 | — | — | Apple M1, 8 GB | | ||
| | python | Windows | CPU | — | — | — | — | | ||
| | python | Windows | NVIDIA GPU | — | — | — | — | | ||
| | cpp | Linux | CPU (OMP) | — | — | — | — | | ||
| | cpp | Linux | NVIDIA GPU (CUDA) | — | — | — | — | | ||
| | cpp | macOS (Apple Silicon) | CPU (OMP) | — | — | — | — | | ||
| | cpp | Windows | CPU (OMP) | — | — | — | — | | ||
| | cpp | Windows | NVIDIA GPU (CUDA) | — | — | — | — | | ||
|
|
||
| Contributions welcome — open a PR filling a row with your k-wave-python version, and (for `cpp` backend) `BINARY_VERSION`. See [`benchmarks/README.md`](../benchmarks/README.md) for the reproducer command. | ||
|
Contributor
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more.
The linked default command runs all 12 benchmark cases, including the large cases excluded from this matrix because workstations may run out of memory. A contributor following this instruction to fill a 64³, 128³, or 256³ cell can therefore spend time on unrelated cases or encounter an out-of-memory failure before the command completes; provide a bounded command that reproduces the listed sizes. |
||
|
|
||
| ## Installation | ||
|
|
||
| Using [uv](https://docs.astral.sh/uv/) (recommended): | ||
|
|
||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The link is labeled
benchmarks/benchmark.py, but it opensbenchmarks/README.md. Readers clicking it to inspect the benchmark implementation are sent to the instructions instead; the instructions already have a separate link below.Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!