One Billion Row Challenge (1BRC)

Link: https://github.com/gunnarmorling/1brc/

The One Billion Row Challenge (1BRC) is a challenge that aims to aggregate one billion rows from a text file using various optimization techniques. The implementation in the provided repository showcases a solution using Modern C++.

Comparison with other implementations (On My Machine):

Implementation	Time
lehuyduc	5.575s
tooomasy	8.413s 7.640s
kajott	10.580s

Key features of the implementation include:

• SIMD instructions for efficient parallel processing.
• Custom hash table with linear probing for fast data lookup.
• Memory mapping technique for efficient file reading.
• Multi-threaded processing to leverage parallelism.
• Usage of std::string_view instead of std::string for improved performance.
• Parsing temperature values as integers instead of floats.

How to run:

mkdir build; cd build; cmake .. && make && ./1brc_cpp <input-filename> <output-result-filename>