Conversation
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
db1e46c to
7b3111a
Compare
CI docs check caught links to pub(crate) constants CODES_OFFSETS_SLOT and CODES_VALIDITY_SLOT in the public FSSTData doc comment. Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk> Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Polar Signals Profiling ResultsLatest Run
Powered by Polar Signals Cloud |
Benchmarks: PolarSignals ProfilingVortex (geomean): 0.920x ➖ datafusion / vortex-file-compressed (0.920x ➖, 2↑ 0↓)
|
File Sizes: PolarSignals Profiling
Totals:
|
Benchmarks: TPC-H SF=1 on NVMEVerdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.999x ➖, 0↑ 0↓)
datafusion / vortex-compact (0.993x ➖, 0↑ 0↓)
datafusion / parquet (1.005x ➖, 0↑ 1↓)
datafusion / arrow (0.993x ➖, 0↑ 1↓)
duckdb / vortex-file-compressed (1.110x ❌, 0↑ 15↓)
duckdb / vortex-compact (1.059x ➖, 0↑ 6↓)
duckdb / parquet (1.061x ➖, 1↑ 4↓)
duckdb / duckdb (1.083x ➖, 0↑ 8↓)
Full attributed analysis
|
File Sizes: TPC-H SF=1 on NVME
Totals:
|
Benchmarks: FineWeb NVMeVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (1.000x ➖, 0↑ 1↓)
datafusion / vortex-compact (1.016x ➖, 0↑ 0↓)
datafusion / parquet (1.010x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed (1.093x ➖, 0↑ 4↓)
duckdb / vortex-compact (1.062x ➖, 0↑ 2↓)
duckdb / parquet (0.994x ➖, 0↑ 0↓)
Full attributed analysis
|
File Sizes: FineWeb NVMe
Totals:
|
Benchmarks: TPC-DS SF=1 on NVMEVerdict: No clear signal (low confidence) datafusion / vortex-file-compressed (0.918x ➖, 32↑ 0↓)
datafusion / vortex-compact (0.896x ✅, 61↑ 3↓)
datafusion / parquet (0.926x ➖, 18↑ 0↓)
duckdb / vortex-file-compressed (0.986x ➖, 14↑ 7↓)
duckdb / vortex-compact (0.899x ✅, 43↑ 5↓)
duckdb / parquet (0.954x ➖, 6↑ 1↓)
duckdb / duckdb (0.967x ➖, 13↑ 8↓)
Full attributed analysis
|
File Sizes: TPC-DS SF=1 on NVME
Totals:
|
🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨Benchmark |
File Sizes: TPC-H SF=10 on NVMENo file size changes detected. |
Benchmarks: TPC-H SF=1 on S3Verdict: No clear signal (environment too noisy confidence) datafusion / vortex-file-compressed (0.928x ➖, 2↑ 0↓)
datafusion / vortex-compact (1.128x ➖, 0↑ 2↓)
datafusion / parquet (0.889x ➖, 4↑ 0↓)
duckdb / vortex-file-compressed (1.038x ➖, 1↑ 0↓)
duckdb / vortex-compact (1.084x ➖, 0↑ 1↓)
duckdb / parquet (1.089x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: FineWeb S3Verdict: No clear signal (low confidence) datafusion / vortex-file-compressed (0.966x ➖, 0↑ 0↓)
datafusion / vortex-compact (0.952x ➖, 0↑ 0↓)
datafusion / parquet (1.018x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed (0.981x ➖, 0↑ 0↓)
duckdb / vortex-compact (0.984x ➖, 0↑ 0↓)
duckdb / parquet (0.992x ➖, 0↑ 0↓)
Full attributed analysis
|
Benchmarks: Random AccessVortex (geomean): 0.916x ➖ unknown / unknown (1.015x ➖, 7↑ 11↓)
|
Benchmarks: Statistical and Population GeneticsVerdict: No clear signal (low confidence) duckdb / vortex-file-compressed (1.045x ➖, 0↑ 1↓)
duckdb / vortex-compact (1.041x ➖, 0↑ 1↓)
duckdb / parquet (1.028x ➖, 0↑ 1↓)
Full attributed analysis
|
File Sizes: Statistical and Population Genetics
Totals:
|
🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨Benchmark |
File Sizes: Clickbench on NVME
Totals:
|
🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨Benchmark |
Benchmarks: CompressionVortex (geomean): 0.998x ➖ unknown / unknown (0.994x ➖, 8↑ 4↓)
|
Merging this PR will degrade performance by 14.72%
Performance Changes
Comparing Footnotes
|
The FSST arrays used to have a children that MUST be a VarBin. However this doesn't work with slots.
Instead the FSST array has a offsets child and a bytes buffer.
This is not a break its backwards compat