I am using Cuda 3.1. Boot your SL in 64 bits, then go to /Developer/GPU Computing/C and 'make'. Everything compiled and all bins in "/Developer/GPU Computing/C/bin/darwin/release" are, e.g.,
volumeRender: Mach-O 64-bit executable x86_64 Doing 'file /usr/local/cuda/lib/*' and all libs there are now universal, e.g.: /usr/local/cuda/lib/libcublas.dylib: Mach-O universal binary with 2 architectures /usr/local/cuda/lib/libcublas.dylib (for architecture x86_64): Mach-O 64-bit dynamically linked shared library x86_64 /usr/local/cuda/lib/libcublas.dylib (for architecture i386): Mach-O dynamically linked shared library i386 So, I believe now that finally Cuda lib and API for Mac are full 64 bits (although 'file /usr/local/cuda/bin/*' are all still i386). Now, after a couple of months, I will give another try to pycuda and pyopencl. Cheers, Alan On Tue, Jul 20, 2010 at 02:59, Andreas Kloeckner <lists@ informa.tiker.net> wrote: > On Mon, 19 Jul 2010 16:22:22 -0700, Paul Northug <pnort...@gmail.com> > wrote: > > I thought with CUDA 3.1, I would be finally able to use a 64-bit CUDA > > with pycuda on Snow Leopard, but maybe it is not so easy. > > Can one of you compile one of Nvidia's driver API examples on 64-bit Mac > and report back whether that works? (if you haven't already done so) > > Andreas > > _______________________________________________ > PyCUDA mailing list > PyCUDA@tiker.net > http://lists.tiker.net/listinfo/pycuda > > -- Alan Wilter S. da Silva, D.Sc. - CCPN Research Associate Department of Biochemistry, University of Cambridge. 80 Tennis Court Road, Cambridge CB2 1GA, UK. >>http://www.bio.cam.ac.uk/~awd28<<
_______________________________________________ PyCUDA mailing list PyCUDA@tiker.net http://lists.tiker.net/listinfo/pycuda