Compression17 chartsencode 2.0× · decode 5.1× faster than Parquet
Write Throughput
#1vortex2.04x1992 MB/s
#2parquet1.00x1134 MB/s
#3lance0.73x761 MB/s
Scan Throughput
#1vortex5.15x51.23 GB/s
#2parquet1.00x9.90 GB/s
#3lance0.95x8.27 GB/s
Geomean throughput ratio to Parquet | Aggregate throughput (higher is better)
Compression Size26 charts14% larger than Parquet
Vs Arrow
#1parquet6.54x
#2vortex5.75x
#3lance3.09x
#4arrow-ipc1.1x
Vs ParquetMinMeanMax
parquet: minimum 1x, mean 1x, maximum 1x
vortex: minimum 0.677x, mean 1.14x, maximum 1.55x
lance: minimum 1.58x, mean 2.11x, maximum 2.74x
arrow-ipc: minimum 3.24x, mean 5.95x, maximum 9.1x
Geometric means of compressed sizes versus Arrow (higher is better) and versus Parquet-zstd (lower is better)
String Encoding21 charts
Write Throughput
#1fsst1.00x222 MB/s
#2onpair-120.64x153 MB/s
#3onpair-160.63x149 MB/s
Scan Throughput
#1onpair-161.23x4.60 GB/s
#2onpair-121.21x4.61 GB/s
#3fsst1.00x3.72 GB/s
Geomean throughput ratio to FSST | Aggregate throughput (higher is better)
String Encoding Size21 charts
Vs Arrow
#1onpair-162.76x
#2onpair-122.76x
#3fsst2.14x
Vs FSSTMinMeanMax
onpair-16: minimum 0.607x, mean 0.777x, maximum 0.9x
onpair-12: minimum 0.6x, mean 0.778x, maximum 0.877x
fsst: minimum 1x, mean 1x, maximum 1x
Geometric means of file sizes versus Arrow (higher is better) and versus FSST (lower is better)
Random Access (NVMe)9 chartshot 102× · cold 32× faster than Parquet
Hot
#1vortex1.00x780.98 us
#2litdata2.40x1.72 ms
#3lance3.08x3.07 ms
#4arrow-ipc4.81x10.49 ms
#5parquet102.39x79.76 ms
Cold
#1vortex1.17x2.67 ms
#2lance1.90x4.56 ms
#3arrow-ipc1.94x10.69 ms
#4litdata20.37x43.73 ms
#5parquet38.09x79.22 ms
Geomean of take time ratio to fastest across every dataset (lower is better)
Random Access (S3)9 chartshot 85× · cold 2.4× faster than Parquet
Hot
#1vortex1.00x20.85 ms
#2lance29.67x173.44 ms
#3parquet84.92x471.47 ms
Cold
#1vortex1.00x266.79 ms
#2lance2.22x582.68 ms
#3parquet2.45x636.39 ms
Geomean of take time ratio to fastest across every dataset (lower is better)
Clickbench43 chartsDuckDB 1.8× · DataFusion 1.6× faster than Parquet
Performance Summary
#1duckdb:vortex1.10x3.66 s
#2datafusion:vortex1.26x3.93 s
#3duckdb:vortex-compact1.31x4.22 s
#4datafusion:vortex-compact1.56x4.94 s
#5datafusion:parquet1.96x5.91 s
#6duckdb:duckdb1.98x5.37 s
#7duckdb:parquet2.02x5.65 s
#8datafusion:lance16.89x47.68 s
Geomean of query time ratio to fastest (lower is better)
TPC-H (NVMe) (SF=1)22 chartsDuckDB 3.7× · DataFusion 3.9× faster than Parquet
Performance Summary
#1duckdb:duckdb1.11x335.77 ms
#2duckdb:vortex1.28x368.82 ms
#3duckdb:vortex-compact1.41x402.76 ms
#4datafusion:vortex1.71x525.48 ms
#5datafusion:vortex-compact1.80x543.17 ms
#6datafusion:arrow3.29x1.08 s
#7duckdb:parquet4.69x1.39 s
#8datafusion:lance6.20x2.01 s
#9datafusion:parquet6.61x1.90 s
Geomean of query time ratio to fastest (lower is better)
TPC-H (S3) (SF=1)22 chartsDuckDB 1.1× faster · DataFusion 1.1× slower vs Parquet
Performance Summary
#1datafusion:parquet1.17x5.59 s
#2datafusion:vortex-compact1.18x5.66 s
#3datafusion:vortex1.29x6.20 s
#4duckdb:vortex-compact1.48x6.98 s
#5datafusion:lance1.54x7.69 s
#6duckdb:vortex1.59x7.57 s
#7duckdb:parquet1.67x7.97 s
Geomean of query time ratio to fastest (lower is better)
TPC-H (NVMe) (SF=10)22 chartsDuckDB 2.5× · DataFusion 2.3× faster than Parquet
Performance Summary
#1duckdb:duckdb1.14x1.02 s
#2duckdb:vortex1.31x1.18 s
#3duckdb:vortex-compact1.48x1.30 s
#4datafusion:vortex1.79x1.82 s
#5datafusion:vortex-compact1.98x1.93 s
#6duckdb:parquet3.22x2.90 s
#7datafusion:parquet4.18x3.66 s
#8datafusion:arrow14.51x16.39 s
#9datafusion:lance20.96x22.43 s
Geomean of query time ratio to fastest (lower is better)
TPC-H (S3) (SF=10)22 chartsDuckDB 1.5× slower · DataFusion ≈ same vs Parquet
Performance Summary
#1datafusion:vortex-compact1.15x11.29 s
#2datafusion:parquet1.24x12.25 s
#3datafusion:vortex1.28x12.55 s
#4duckdb:parquet1.40x13.35 s
#5duckdb:vortex-compact2.04x20.07 s
#6duckdb:vortex2.12x20.91 s
#7datafusion:lance3.75x43.02 s
Geomean of query time ratio to fastest (lower is better)
TPC-H (NVMe) (SF=100)22 chartsDuckDB 1.7× · DataFusion 1.8× faster than Parquet
Performance Summary
#1duckdb:vortex1.13x7.23 s
#2datafusion:vortex1.66x13.94 s
#3duckdb:parquet1.89x11.34 s
#4datafusion:parquet3.06x19.83 s
Geomean of query time ratio to fastest (lower is better)
TPC-H (S3) (SF=100)22 chartsDuckDB 4.0× slower · DataFusion ≈ same vs Parquet
Performance Summary
#1duckdb:parquet1.06x36.60 s
#2datafusion:parquet1.34x48.56 s
#3datafusion:vortex1.38x50.05 s
#4duckdb:vortex4.22x167.58 s
Geomean of query time ratio to fastest (lower is better)
TPC-DS (NVMe) (SF=1)99 chartsDuckDB 1.3× · DataFusion 1.5× faster than Parquet
Performance Summary
#1duckdb:duckdb1.04x1.70 s
#2duckdb:vortex1.39x2.15 s
#3duckdb:vortex-compact1.63x2.47 s
#4duckdb:parquet1.83x2.82 s
#5datafusion:vortex2.23x4.28 s
#6datafusion:vortex-compact2.42x4.77 s
#7datafusion:parquet3.33x6.04 s
Geomean of query time ratio to fastest (lower is better)
TPC-DS (NVMe) (SF=10)99 chartsDuckDB 1.6× · DataFusion 1.4× faster than Parquet
Performance Summary
#1duckdb:vortex1.09x6.26 s
#2datafusion:vortex1.53x22.92 s
#3duckdb:parquet1.78x8.84 s
#4datafusion:parquet2.08x14.05 s
Geomean of query time ratio to fastest (lower is better)
TPC-DS (NVMe) (SF=100)99 chartsDuckDB ≈ same · DataFusion 1.3× faster vs Parquet
Performance Summary
#1duckdb:parquet1.20x27.08 s
#2duckdb:vortex1.21x29.39 s
#3datafusion:vortex1.78x151.64 s
#4datafusion:parquet2.37x148.10 s
Geomean of query time ratio to fastest (lower is better)
TPC-DS (S3) (SF=10)99 chartsDuckDB ≈ same · DataFusion 1.2× slower vs Parquet
Performance Summary
#1datafusion:parquet1.38x69.03 s
#2duckdb:parquet1.40x61.18 s
#3duckdb:vortex1.44x74.59 s
#4datafusion:vortex1.62x88.67 s
Geomean of query time ratio to fastest (lower is better)
TPC-DS (S3) (SF=100)99 chartsDuckDB 4.8× slower · DataFusion 1.7× faster vs Parquet
Performance Summary
#1datafusion:vortex1.25x256.91 s
#2duckdb:parquet1.28x149.62 s
#3datafusion:parquet2.08x342.81 s
#4duckdb:vortex6.20x965.41 s
Geomean of query time ratio to fastest (lower is better)
Statistical and Population Genetics11 chartsDuckDB 4.2× · DataFusion 3.4× faster than Parquet
Performance Summary
#1datafusion:vortex1.14x1.50 s
#2datafusion:vortex-compact1.23x1.68 s
#3duckdb:vortex2.30x2.49 s
#4duckdb:vortex-compact2.44x2.56 s
#5datafusion:parquet3.85x3.81 s
#6duckdb:parquet9.65x6.72 s
Geomean of query time ratio to fastest (lower is better)
PolarSignals Profiling10 chartsVortex only
Performance Summary
#1datafusion:vortex1.00x375.89 ms
Geomean of query time ratio to fastest (lower is better)
fineweb [nvme]9 chartsDuckDB 5.2× · DataFusion 3.3× faster than Parquet
Performance Summary
#1duckdb:vortex1.01x210.42 ms
#2datafusion:vortex1.34x226.41 ms
#3duckdb:vortex-compact3.18x762.22 ms
#4datafusion:parquet4.44x629.72 ms
#5duckdb:parquet5.27x788.57 ms
#6datafusion:vortex-compact5.81x1.11 s
Geomean of query time ratio to fastest (lower is better)
fineweb [s3]9 chartsDuckDB 1.3× · DataFusion 2.1× faster than Parquet
Performance Summary
#1datafusion:vortex-compact1.08x3.50 s
#2datafusion:vortex1.09x3.60 s
#3duckdb:vortex-compact1.52x5.27 s
#4duckdb:vortex1.56x5.53 s
#5duckdb:parquet1.99x5.52 s
#6datafusion:parquet2.32x6.11 s
Geomean of query time ratio to fastest (lower is better)
appian [nvme]8 chartsDuckDB ≈ same · DataFusion 1.3× faster vs Parquet
Performance Summary
#1duckdb:duckdb1.13x1.62 s
#2duckdb:parquet1.24x1.67 s
#3duckdb:vortex1.30x1.80 s
#4duckdb:vortex-compact1.32x1.85 s
#5datafusion:vortex1.87x10.99 s
#6datafusion:vortex-compact1.89x11.03 s
#7datafusion:parquet2.37x11.42 s
Geomean of query time ratio to fastest (lower is better)
clickbench-sorted [nvme]10 chartsDuckDB 2.2× · DataFusion 2.7× faster than Parquet
Performance Summary
#1duckdb:vortex1.01x174.57 ms
#2duckdb:vortex-compact1.21x214.73 ms
#3datafusion:vortex1.31x319.95 ms
#4datafusion:vortex-compact1.63x375.11 ms
#5duckdb:duckdb2.07x343.66 ms
#6duckdb:parquet2.19x442.48 ms
#7datafusion:parquet3.56x1.28 s
#8datafusion:lance75.50x19.89 s
Geomean of query time ratio to fastest (lower is better)
vortex [nvme]3 chartsDuckDB 1.4× faster · DataFusion ≈ same vs Parquet
Performance Summary
#1datafusion:parquet1.18x491.18 ms
#2datafusion:vortex1.21x433.11 ms
#3duckdb:vortex1.52x547.34 ms
#4duckdb:parquet2.11x780.16 ms
Geomean of query time ratio to fastest (lower is better)