It is poorly optimized for 32bits - some benchmarks: http://encode.ru/threads/2221-LZFSE-New-Apple-Data-Compressi...