Reference library
Parallel Random Numbers: As Easy as 1, 2, 3
Paper summary
Introduces counter-based pseudorandom generators, including Philox, that map keys and counters to samples without sequential mutable stream state. Statistical and performance evaluations target massively parallel computation.
Relevance to the project
Directly relevant to schedule-independent initialization, mutation, and stochastic environment events on future GPUs. Logical event identities can key draws while intrinsic constant-input neural tests remain noise-free.
RNG reproducibility alone does not guarantee floating-point or phase-order reproducibility.
Source record
- Reference ID
- parallel-random-numbers-2011
- Archived version
- Author-hosted ACM conference paper; SC11 copyright/citation on first page
- Summary provenance
- reference-library/papers/parallel-random-numbers-2011/metadata.json · summary
- Review status
- initial orientation summary; not a critical full-paper review
- Reading scope
- Selected abstract/opening and available closing discussion/conclusion passages in the preserved text extract; broader project notes provide context. No reproduction of experiments, exhaustive methods review, or equation/figure audit.