EXPERIMENT RECORD
WEB-EXP-008
Hybrid Spatial Index Boundary Query Benchmark
Hardware benchmark of a boundary query using a Mulberry32 pseudo-random sequence with seed 424242 and N=100,000.
Reported result
| candidate_count | 100000 |
|---|---|
| boundary_query_ms | 0.481 |
| operating_window | K ∈ [32, 256] |
Interpretation
The reported result supports the hypothesis that sub-tree promotion can avoid candidate-array scanning bottlenecks under the tested conditions.
Limitations
- The result is specific to the reported implementation and test conditions.
- A benchmark result is not a universal performance guarantee.
- The operating window K ∈ [32, 256] is reported as an observed benchmark window, not a theorem.
- True raw execution logs must be supplied separately if external reviewers are to reproduce the exact run trace.
Published result exports
These files contain the benchmark values published in the registry. They are result exports, not raw execution logs.
Raw logs are not represented here unless the original execution-log files are supplied.