Parallel Random Number Generators

  • Ronald T. Kneusel


It is often the case that many separate threads or processes require independent streams of pseudorandom numbers. This chapter examines five methods for generating such streams: a pseudorandom number server process, separate per stream generators, partitioning a single stream into non-overlapping segments, random seeding that relies on the small likelihood of randomly picking overlapping streams, and merging two randomly initialized generator outputs. Implementations for certain generators will be developed and output streams tested.


  1. 1.
  2. 2.
    Marsaglia, George; Sullivan, Stephen. “Technical correspondence”. Communications of the ACM. 36 (7) (1993): 105–110.CrossRefGoogle Scholar
  3. 3.
    Marsaglia, George. “Xorshift rngs.” Journal of Statistical Software 8.14 (2003): 1–6.CrossRefGoogle Scholar
  4. 4.
    Kneusel, Ronald T. “Curve-Fitting on Graphics Processors Using Particle Swarm Optimization.” International Journal of Computational Intelligence Systems 7, no. 2 (2014): 213–224.Google Scholar
  5. 5.
    Nguyen, Hubert. Gpu gems 3. Addison-Wesley Professional, 2007.Google Scholar
  6. 6.
    L’Ecuyer, Pierre, David Munger, Boris Oreshkin, and Richard Simard. “Random numbers for parallel computers: Requirements and methods, with emphasis on gpus.” Mathematics and Computers in Simulation 135 (2017): 3–17.MathSciNetCrossRefGoogle Scholar
  7. 7.
    L’ecuyer, Pierre. “Tables of linear congruential generators of different sizes and good lattice structure.” Mathematics of Computation of the American Mathematical Society 68, no. 225 (1999): 249–260.MathSciNetCrossRefzbMATHGoogle Scholar
  8. 8.
    Fog, Agner. “Pseudo-Random Number Generators for Vector Processors and Multicore Systems.” Journal of Modern Applied Statistical Methods 14, no. 1 (2015): 308–334.CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Ronald T. Kneusel
    • 1
  1. 1.ThorntonUSA

Personalised recommendations