LZ4 is a very fast lossless compression algorithm, providing compression speed > 500 MB/s per core, scalable with multi-cores CPU. It also features an extremely fast decoder, with speed in multiple GB/s per core, typically reaching RAM speed limits on multi-core systems.
LZ4 is available as a C Open Source project, hosted on Github, under a BSD license. A list of ports to other language (Java, C#, Python, etc.) is also maintained on this page.
LZ4 compression format is detailed in a dedicated post.
LZ4 project is currently hosted on Github.
Pre-compiled binaries for Windows are proposed in the release section.
Compression level comparison :
In-memory benchmark using lzbench, on the Silesia Corpus
Benchmark platform : Core i7-3930K @ 4.5GHz, Linux 64-bits, gcc v6.2
|Compression / Thread||Ratio||Speed||Decoding|
|LZ4 Fast 8||1.799||911 MB/s||3.36 GB/s|
|LZ4 default||2.101||625 MB/s||3.22 GB/s|
|LZ4 HC 9||2.720||34 MB/s||3.24 GB/s|
You can provide comments and ask questions on the LZ4 Forum Group or the LZ4 issue board .
Should you wish to provide a token of your appreciation, the following link will allow you to donate to the LZ4 project :