Skip to content

Latest commit

 

History

History
37 lines (25 loc) · 1.56 KB

BENCHMARK.md

File metadata and controls

37 lines (25 loc) · 1.56 KB
Raw

Benchmarks

Benchmarks are set up for GeoJSON serialization and deserialization.

Running Benchmarks

./gradlew :geojson:benchmark

This will run benchmarks on the JVM, NodeJS, and Kotlin/Native.

These benchmarks measure the time taken to serialize and deserialize a FeatureCollection containing 15,000 randomly generated features. See GeoJsonBenchmark.kt for details.

Results

All measurements are in ms/op (milliseconds per operation). Lower score is better.

Target Serialization (fast) Serialization (kotlinx) Deserialization
JVM 71.450 ± 2.566 250.776 ± 7.869 84.444 ± 1.968
JS 178.458 ± 5.962 986.504 ± 24.773 383.339 ± 15.627
Native (linuxX64) 326.828 ± 32.121 1073.703 ± 168.403 652.600 ± 117.655

Run on Ubuntu 20.04 (WSL2). 32GB RAM, 3.60 GHz 8-core Intel Core i7-9700k

Fast vs. Kotlinx

There are two serialization implementations available in Spatial-K. The "kotlinx" implementation uses kotlinx.serialization to encode the object to a string and is fully compatible with other kotlinx.serialization models, but is slower. The "fast" implementation encodes an object directly to a string using string interpolation and isn't as versatile, but is much faster.

Deserialization is done only using kotlinx.serialization.