Advection, weak scaling
The problem grows with the number of MPI ranks.
Tags: cpu, weak-scaling
| run | file | recorded | ranks | wall time | vs baseline | page |
|---|---|---|---|---|---|---|
| np1 | np1.h5 | 2026-10-17 12:00 UTC | 1 | 124 ms | baseline | run page |
| np2 | np2.h5 | 2026-10-17 12:00 UTC | 2 | 153 ms | +23.5% | run page |
| np4 | np4.h5 | 2026-10-17 12:00 UTC | 4 | 161 ms | +29.4% | run page |
| np8 | np8.h5 | 2026-10-17 12:00 UTC | 8 | 187 ms | +50.6% | run page |
Each run's page is linked on the right.
What changed
Section titled “What changed”Durations
Section titled “Durations”Charts
Section titled “Charts”Region durations
Grouped bars compare each region's mean call duration across the profiled runs.
Weak scaling
Each line is one region's weak-scaling efficiency: its mean call duration on the run with the fewest MPI ranks divided by its mean call duration on each run. The dashed line at 1 is ideal: the same time per call however large the run; a region below it loses time to costs that grow with the run, such as communication or load imbalance. The 8 regions with the most time on that first run are shown. Valid for a weak-scaling study: the problem grows with the run, so every run has the same problem size per rank or core.
Rank imbalance
Each line compares a region's total duration by rank; its dashed line marks the mean across ranks. Points far from the mean identify stragglers.
Rank heatmap
This heatmap uses exclusive timings. Exclusive duration is the time spent in a region itself, excluding time spent in nested child regions; this prevents the enclosing session region from dominating the heatmap.
Metadata
Section titled “Metadata”| commit | 0000000 |
|---|---|
| system | synthetic |