Abstract
Large scale Monte Carlo applications need a good pseudo-random number generator
capable of utilizing both the vector processing capabilities and multiprocessing
capabilities of modern computers in order to get the maximum performance. The
requirements for such a generator are discussed. New ways of avoiding overlapping
subsequences by combining two generators are proposed. Some fundamental
philosophical problems in proving independence of random streams are discussed.
Remedies for hitherto ignored quantization errors are offered. An open source C++
implementation is provided for a generator that meets these needs.
| Original language | English |
|---|---|
| Article number | 23 |
| Journal | Journal of Modern Applied Statistical Methods |
| Volume | 14 |
| Issue number | 1 |
| Pages (from-to) | 308-334 |
| ISSN | 1538-9472 |
| Publication status | Published - 2015 |
Keywords
- Random number generation
- SIMD
- Vector processors
- Multiprocessors
- Parallel generation
- Combination of generators
- Quantization errors
- Theoretical proofs
- Philosophy of science