I have CUDA running smoothly on my desktop, a Windows XP 32 bit installation.
Teh code which reads images into my program has to be compiled in 32 bit mode.
This is nto a problme on WinXP as its the only option.
Well, I now have a server system that run Windows Server 2008 (64 bit OS) with
an S1070 attached. The 64 bit OS driver has been installed for the S1070.
VS2008 incorporates two C compilers, 32 bit and 64 bit.
On this system I eventually found that I could compile my CUDA code
by setting the -m32 flag in the Additional options box of the CUDA Build Rule.
Will the 32 bit executable interoperate correctly with the CUDA 64 bit driver installed.