Note
For an interpretation of the benchmark results, read the Performance section in the main README.md file.
Tested on a MacBook Pro 16" 2023 with an M3 Pro Chip and 36 GB of RAM.
Tested on a MacBook Pro 16" 2023 with an M3 Pro Chip and 36 GB of RAM.
Reads the input file with a BufReader and parses the read bytes with Actson.
Uses Tokio to asynchronously read the input file. Parses it with Actson.
Uses two Tokio tasks: one that reads the file asynchronously and one that parses the read bytes with Actson. The tasks communicate with each other through a channel. This benchmark demonstrates how reading and parsing can happen concurrently.
Parses the JSON file with Serde JSON into a Value.
Deserializes the JSON file with Serde JSON into a struct.
Uses a Serde JSON with a custom deserializer to avoid having to load the whole file into memory (see example on the Serde website).
The benchmarks automatically read every JSON file in the data directory.
To obtain the results above, we've used GeoJSON files from Microsoft Maps containing building footprints for the United States: https://github.com/microsoft/USBuildingFootprints
These files have been released under the Open Data Commons Open Database License (ODbL).
You may use any file from this data set but to reproduce the results above, download the following files and extract them to the data directory:
| State or district | Number of buildings | Unzipped size |
|---|---|---|
| Alaska | 111,042 | 31.46 MB |
| California | 11,542,912 | 3.60 GB |
| Florida | 7,263,195 | 2.16 GB |
| Georgia | 3,981,792 | 1.12 GB |
| Idaho | 942,132 | 272.03 MB |
| Minnesota | 2,914,016 | 799.10 MB |
| North Carolina | 4,678,064 | 1.31 GB |
| Oregon | 1,873,786 | 572.46 MB |
| Vermont | 351,266 | 92.19 MB |
For convenience, here's a script that downloads and extracts all files into the current working directory:
#!/usr/bin/env sh
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/Alaska.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/California.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/Florida.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/Georgia.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/Idaho.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/Minnesota.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/NorthCarolina.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/Oregon.geojson.zip
wget https://usbuildingdata.blob.core.windows.net/usbuildings-v2/Vermont.geojson.zip
unzip Alaska.geojson.zip
unzip California.geojson.zip
unzip Florida.geojson.zip
unzip Georgia.geojson.zip
unzip Idaho.geojson.zip
unzip Minnesota.geojson.zip
unzip NorthCarolina.geojson.zip
unzip Oregon.geojson.zip
unzip Vermont.geojson.zipIn order to run the benchmarks, you need hyperfine installed on your system. Download it here if you do not have it already.
Execute the following script to run all benchmarks:
./run-all.sh
The command will run for about 40-45 minutes, depending on your system. It will write the result to results.json. Open the provided Jupyter Notebook results.ipynb to visualize them.