feat(vortex-bench): infra of SpatialBench on DuckDB, plain binary#8598
feat(vortex-bench): infra of SpatialBench on DuckDB, plain binary#8598HarukiMoriarty wants to merge 4 commits into
Conversation
HarukiMoriarty
changed the title
| threads: Option<usize>, | ||
| ) -> Result<(Database, Connection)> { | ||
| let db = match path { | ||
| Some(path) => Database::open(path), |
There was a problem hiding this comment.
Seems like these changes are already in develop, can you rebase?
There was a problem hiding this comment.
ok, I will rebase it.
Signed-off-by: Nemo Yu <zyu379@wisc.edu>
HarukiMoriarty
force-pushed
the
nemo/spatial-wkb
branch
from
f1bff93 to
dcdae4f
Compare
Merging this PR will degrade performance by 14.66%
|
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ❌ | Simulation | slice_empty_vortex |
339.4 ns | 397.8 ns | -14.66% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing nemo/spatial-wkb (c1635cb) with develop (3451cb0)
Footnotes
-
4 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩
|
Fixed, wire to |
3 participants
Summary
Adds a SpatialBench benchmark that runs on DuckDB across two storage lanes, Vortex binary BLOB and Parquet binary BLOB.
Benchmark results
Local, warm, median of 5 (ms).
Takeaways
Vortex wins the scan/filter-bound queries (Q1–Q4, Q6) by up to 3.5×. Q5/Q7/Q8/Q9 are compute-bound (DuckDB spatial join dominates), so the formats converge. Q10–Q12 time out on both, because the unindexed spatial join is the wall, independent of storage format.