On Mon, Jan 30, 2012 at 8:21 PM, Andreas Kloeckner <li...@informa.tiker.net> wrote: > On Sat, 28 Jan 2012 18:21:29 -0500, Thomas Wiecki <thomas_wie...@brown.edu> > wrote: >> I am currently revisiting this but having some problems with the >> random number generator. >> >> generator.generators_per_block is 512 on my card, so I initialize 512 >> generators, but I see that some of them don't produce random numbers >> when sampling from them. I notice that in some subtle ways (mainly >> that the distribution is not correct or all numbers are the same) if I >> sample from more than 300-350 generators (always the last ones are >> affected), but it's fine when using e.g. 128. So it seems I can only >> use a smaller number of generators than what the card says I should be >> able to use. >> >> Any idea on why that might be or how to investigate this further? > > Mysterious. What generator is this using? XORWOW? Tomasz, any ideas? > > Andreas
Currently revisiting this with updated everything. So far I can't reproduce my earlier odd behavior (after a certain size the random numbers generated did not look random anymore). One thing I did notice was that before generators_per_block returned 512, while now it returns 384. Interestingly that seems in the same ball-park as the number I found empirically before (see above). Was that a (Py)Cuda/curand bug perhaps that got fixed in the interim? On Fri, Dec 23, 2011 at 10:21 PM, Andreas Kloeckner <li...@informa.tiker.net> wrote: > Not for 2D arrays, but for 1D-arrays-viewed-as-2D-arrays--that's > generator.fill_uniform() for example. There was a doc bug that kept > these from showing up, fixed. What is a 1D-arrays-viewed-as-2D-arrays? Is the key the call to prepared_async_call()? Does this method take care of calling the function the appropriate number of times if the array to fill is bigger than generators_per_block? Would this method help me if I had my own random process that I wanted to use on arbitrary large array sizes? Thomas _______________________________________________ PyCUDA mailing list PyCUDA@tiker.net http://lists.tiker.net/listinfo/pycuda