SIMD-Oriented Fast Mersenne Twister: a 128-bit Pseudorandom Number Generator

By Mutsuo Saito, Makoto Matsumoto. 2008

301 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

State-of-the-art in heterogeneous computing (2010, methodology): “A variant of the Mersenne Twister is the SIMD-oriented Fast Mersenne Twister [153], which utilizes the 128-bit operations provided by SSE2 and AltiVec, and roughly doubles the performance.”
Fast Random Integer Generation in an Interval (2018, background): “While generating a whole new 64b pseudo-random number can take as little as a handful of cycles [33], transforming it into an integer in an interval ([0, s ) for s ∈ [0, 2(64))) without bias can take an order of magnitude longer when using division operations.”
Detailed Simulations of Cell Biology with Smoldyn 2.1 (2010, methodology): “This core is linked to the OpenGL library for graphics, the libtiff library for saving tiff format images, the libmoleculizer library for rule-based reaction network generation [32], the POSIX library for threaded operation, and the SIMD-oriented fast Mersenne Twister library [67] for random number ”
Handbook of Monte Carlo Methods (2011, background): “[35], which correct some weaknesses in MT19937, and the SIMD-oriented fast Mersenne twister [38], which is significantly faster than the standard Mersenne twister, has better equidistribution properties, and recovers faster from states with many 0s.”