Xorshift1024*, Xorshift1024+, Xorshift128+ and Xoroshiro128+ Fail Statistical Tests for Linearity

Abstract L’Ecuyer & Simard’s Big Crush statistical test suite has revealed statistical flaws in many popular random number generators including Marsaglia’s Xorshift generators. Vigna recently proposed some 64-bit variations on the Xorshift scheme that are further scrambled (i.e., xorshift1024*, xorshift1024+, xorshift128+, xoroshiro128+). Unlike their unscrambled counterparts, they pass Big Crush when interleaving blocks of 32 bits for each 64-bit word (most significant, least significant, most significant, least significant, etc.). We report that these scrambled generators systematically fail Big Crush – specifically the linear-complexity and matrix-rank tests that detect linearity – when taking the 32 lowest-order bits in reverse order from each 64-bit word.