1 Billion Record Challenge in Perl
Original challenge for Java: https://github.com/gunnarmorling/1brc
To get the full input file, clone that repo according the instructions and run the commands to generate it. It's not included in this repo.
I have included the first 1M rows in the zipped file "test.zip".
My machine specs:
Model name: Intel(R) Core(TM) i7-5557U CPU @ 3.10GHz
Output of free -h:
total used free shared buff/cache available
Mem: 15Gi 1,4Gi 195Mi 472Mi 13Gi 13Gi
Swap: 2,0Gi 1,9Gi 148Mi
Java reference implementation in the original repo runs in around 3m
Using cat to write the file to /dev/null takes ~18s.