LZ4 v1.10 Introduces Multi-Threading Support For Major Compression Speedups

Yann Collet released LZ4 v1.10 today as a major update to this extremely fast compression algorithm. Most significant with LZ4 1.10 is adding multi-threaded compression support for much faster performance with today's modern multi-core processors.

LZ4 1.10 has been dubbed the "multi-cores edition" with this version adding multi-threading support to help speed-up compression now that modern I/O storage with NVMe is so much faster there's a real need to make compressing data even faster.

Benchmarks by Yann Collet show LZ4 1.10 with multi-threading being 5.4~8.0x faster with various data and across Intel / AMD / Apple processors on Windows / macOS / Linux. LZ4 compression with multi-threading on any modern CPUs should now be much, much faster than prior versions. LZ4 decompression can also be ~60%+ faster with overlapping decompression but not as important as the compression speed-ups.

LZ4 1.10 also promotes its dictionary compression from being "experimental" to a fully supported feature. There's also a new "Level 2" compression mode for LZ4, build system improvements, various CLI fixes, and a variety of other optimizations and code improvements.

Downloads and more details on this big LZ4 v1.10 update via GitHub. I'll have out some new LZ4 benchmarks soon across different processors in looking at the multi-threaded compression performance gains.