[PyCUDA] how to detect maximum array i can multiply?

[arief nur andono]

hi...i'm upgrade my operating system to ubuntu 10.04 so i move to cuda
version 3.2 RC with pycuda 0.94.1

but when i tried to compile some example i got some errors that i didn't get
before...

i always get these errors even from examples at
wiki.tiker.net/PyCuda/Examples

these are the errors


/usr/local/lib/python2.6/dist-packages/pycuda-0.94.1-py2.6-linux-i686.egg/pycuda/compiler.py:112:
UserWarning: The CUDA compiler suceeded, but said the following:
ptxas /tmp/tmpxft_0000452b_00000000-2_kernel.ptx, line 86; warning : Double
is not supported. Demoting to float

  +stdout+stderr)
Traceback (most recent call last):
  File "mandelfun.py", line 67, in <module>
    block=(500,1,1), grid = (50000,1))
  File
"/usr/local/lib/python2.6/dist-packages/pycuda-0.94.1-py2.6-linux-i686.egg/pycuda/driver.py",
line 159, in function_call
    handlers = func.param_set(*args)
  File
"/usr/local/lib/python2.6/dist-packages/pycuda-0.94.1-py2.6-linux-i686.egg/pycuda/driver.py",
line 119, in function_param_set
    arg_data.append(int(arg.get_device_alloc()))
  File
"/usr/local/lib/python2.6/dist-packages/pycuda-0.94.1-py2.6-linux-i686.egg/pycuda/driver.py",
line 49, in get_device_alloc
    self.dev_alloc = mem_alloc_like(self.array)
  File
"/usr/local/lib/python2.6/dist-packages/pycuda-0.94.1-py2.6-linux-i686.egg/pycuda/driver.py",
line 328, in mem_alloc_like
    return mem_alloc(ary.nbytes)
pycuda._driver.MemoryError: cuMemAlloc failed: out of memory

could you share to solve this??
regards
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tiker.net/pipermail/pycuda/attachments/20101002/09b3fe1e/attachment.html>