Re: parallel random numbers generator

On 2008-05-04 19:19:47 -0300, Crni Gorac <cgorac@xxxxxxxxx> said:

OK, let me try to be somewhat more specific: Particular needs for this
simulation are alike, I guess, to needs of most of MC based
simulations to be implemented on CUDA, which means the demands from
RNG are: to generate many large sequences of random numbers in
parallel, to have these sequences manifesting low correlation between
each other and being generally "good" in other statistical regards,
then to have small state (Mersenne Twister need 624 numbers of state,
this is way too big) in order to be able to keep it in CUDA shared
memory (otherwise the performance will suffer heavily). Seems like
good choice may be to use some of RNGs with small state for each
thread of execution, but initialized with different seeds (generated
by another RNG). But again - there are way too much options available
for both "main" and "seeding" RNG, so the question is: which ones to
choose, and how to measure the quality of the results?

If it has a small state then it will have a short cycle. If you want
large sequences then you seem to want a long cycle. Having several
points active on a short cycle is not compatible with independence
of long sequences. You seem to keep asking for self contradictory or
mutually incompatible properties.




.

Relevant Pages

  • Re: Global Warming
    ... akin to Plato's Allegory of the Cave. ... this simulation, I would expect exactly 250 heads and 250 tails. ... RNG, rather than a program of nature, like a flaw in mathematics. ...
    (rec.gambling.poker)
  • Re: Toaster to Generate Random Numbers
    ... Suppose you have an RNG that chooses a random integer from 1 to 10. ... keep the past 10 days of newspapers, and use this number to select which of ... them to digest to seed your random number sequence. ... different sequences. ...
    (comp.security.misc)
  • Re: Toaster to Generate Random Numbers
    ... ]Suppose you have an RNG that chooses a random integer from 1 to 10. ... ]keep the past 10 days of newspapers, and use this number to select which of ... ]them to digest to seed your random number sequence. ... ]different sequences. ...
    (comp.security.misc)
  • Re: parallel random numbers generator
    ... simulation are alike, I guess, to needs of most of MC based ... to have these sequences manifesting low correlation between ... by another RNG). ...
    (sci.math.num-analysis)
  • Re: Finding independence seq. of Random numbers
    ... and each sequences has to be independent. ... You could use alternative types of RNG: ... However, using three different seeds will generally be independent, ... Because you need to restart sequences many cryptographic RNGs will not ...
    (comp.lang.java.programmer)