Conversation
Signed-off-by: Adam Gutglick <adam@spiraldb.com>
Benchmarks: Statistical and Population GeneticsVerdict: No clear signal (low confidence) duckdb / vortex-file-compressed (0.866x ✅, 3↑ 0↓)
duckdb / vortex-compact (0.948x ➖, 0↑ 0↓)
duckdb / parquet (0.953x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: TPC-H SF=1 on NVMEVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (1.007x ➖, 0↑ 0↓)
datafusion / vortex-compact (1.012x ➖, 0↑ 0↓)
datafusion / parquet (1.015x ➖, 0↑ 1↓)
datafusion / arrow (1.053x ➖, 0↑ 4↓)
duckdb / vortex-file-compressed (1.003x ➖, 0↑ 0↓)
duckdb / vortex-compact (1.003x ➖, 0↑ 0↓)
duckdb / parquet (0.985x ➖, 1↑ 0↓)
duckdb / duckdb (1.008x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: TPC-H SF=1 on S3Verdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.949x ➖, 2↑ 0↓)
datafusion / vortex-compact (1.130x ➖, 1↑ 4↓)
datafusion / parquet (0.998x ➖, 3↑ 3↓)
duckdb / vortex-file-compressed (1.035x ➖, 0↑ 0↓)
duckdb / vortex-compact (1.011x ➖, 0↑ 0↓)
duckdb / parquet (1.078x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: FineWeb NVMeVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (1.024x ➖, 0↑ 1↓)
datafusion / vortex-compact (0.998x ➖, 0↑ 0↓)
datafusion / parquet (1.000x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed (0.988x ➖, 1↑ 1↓)
duckdb / vortex-compact (1.017x ➖, 0↑ 1↓)
duckdb / parquet (1.008x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: TPC-H SF=10 on S3Verdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.966x ➖, 2↑ 1↓)
datafusion / vortex-compact (1.012x ➖, 0↑ 1↓)
datafusion / parquet (1.095x ➖, 0↑ 3↓)
duckdb / vortex-file-compressed (0.963x ➖, 0↑ 0↓)
duckdb / vortex-compact (0.966x ➖, 0↑ 0↓)
duckdb / parquet (1.023x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: FineWeb S3Verdict: No clear signal (low confidence) datafusion / vortex-file-compressed (0.959x ➖, 0↑ 0↓)
datafusion / vortex-compact (1.027x ➖, 0↑ 0↓)
datafusion / parquet (1.064x ➖, 0↑ 1↓)
duckdb / vortex-file-compressed (1.058x ➖, 0↑ 0↓)
duckdb / vortex-compact (1.052x ➖, 0↑ 0↓)
duckdb / parquet (1.045x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: PolarSignals ProfilingVortex (geomean): 1.131x ❌ datafusion / vortex-file-compressed (1.131x ❌, 0↑ 7↓)
|
Benchmarks: TPC-H SF=10 on NVMEVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (1.162x ❌, 0↑ 18↓)
datafusion / vortex-compact (1.135x ❌, 0↑ 18↓)
datafusion / parquet (1.128x ❌, 0↑ 16↓)
datafusion / arrow (1.164x ❌, 0↑ 20↓)
duckdb / vortex-file-compressed (1.139x ❌, 0↑ 17↓)
duckdb / vortex-compact (1.111x ❌, 0↑ 17↓)
duckdb / parquet (1.040x ➖, 0↑ 2↓)
duckdb / duckdb (1.070x ➖, 0↑ 5↓)
Full attributed analysis
|
Benchmarks: Clickbench on NVMEVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (0.993x ➖, 1↑ 1↓)
datafusion / parquet (0.994x ➖, 1↑ 0↓)
duckdb / vortex-file-compressed (0.967x ➖, 6↑ 2↓)
duckdb / parquet (0.989x ➖, 1↑ 0↓)
duckdb / duckdb (0.996x ➖, 0↑ 1↓)
Full attributed analysis
|
Benchmarks: TPC-DS SF=1 on NVMEVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (1.002x ➖, 0↑ 2↓)
datafusion / vortex-compact (0.995x ➖, 0↑ 0↓)
datafusion / parquet (0.996x ➖, 1↑ 1↓)
duckdb / vortex-file-compressed (1.001x ➖, 2↑ 2↓)
duckdb / vortex-compact (1.001x ➖, 0↑ 0↓)
duckdb / parquet (1.001x ➖, 1↑ 2↓)
duckdb / duckdb (1.004x ➖, 0↑ 2↓)
Full attributed analysis
|
Signed-off-by: Adam Gutglick <adam@spiraldb.com>
Polar Signals Profiling ResultsLatest Run
Previous Runs (4)
Powered by Polar Signals Cloud |
File Sizes: PolarSignals ProfilingFile Size Changes (1 files changed, +0.1% overall, 1↑ 0↓)
Totals:
|
File Sizes: FineWeb NVMeFile Size Changes (2 files changed, +0.0% overall, 2↑ 0↓)
Totals:
|
File Sizes: TPC-H SF=1 on NVMEFile Size Changes (18 files changed, +0.2% overall, 18↑ 0↓)
Totals:
|
File Sizes: TPC-DS SF=1 on NVMEFile Size Changes (48 files changed, +0.8% overall, 48↑ 0↓)
Totals:
|
File Sizes: Statistical and Population GeneticsFile Size Changes (2 files changed, +7.1% overall, 2↑ 0↓)
Totals:
|
File Sizes: TPC-H SF=10 on NVMEFile Size Changes (48 files changed, +0.2% overall, 48↑ 0↓)
Totals:
|
Signed-off-by: Adam Gutglick <adam@spiraldb.com>
7ca7c99 to
294b6a8
Compare
Signed-off-by: Adam Gutglick <adam@spiraldb.com>
File Sizes: Clickbench on NVMEFile Size Changes (201 files changed, +0.7% overall, 200↑ 1↓)
Totals:
|
Merging this PR will improve performance by 27.63%
Performance Changes
Comparing Footnotes
|
vx-bench to run benchmarks in CI
| --opt scale-factor=1.0 | ||
|
|
||
| uv run --project bench-orchestrator vx-bench run tpch \ | ||
| --targets-json '[{"engine":"datafusion","format":"parquet"},{"engine":"duckdb","format":"vortex"}]' \ |
There was a problem hiding this comment.
should we put the opt in this json?
There was a problem hiding this comment.
I was thinking about making this some sort of manifest format that we can just have different version of, WDYT?
There was a problem hiding this comment.
ye I think we should have this. happy to move out of this one
There was a problem hiding this comment.
I feel like this is already pretty big, lets merge it and I'll do a follow-up today/tomorrow
Summary
Use the same
vx-benchtool we want to use locally to run the benchmarks in CI, removing bespoke bash argument parsing logic, generating data with it.