I have installed a guest Ubuntu 10.04 on a vmware player. The host OS is Windows 7 Pro, 32 bit. The computer has a CUDA-enabled graphical card.
The question is: can I access use CUDA from within the virtual machine? this is because the Theano package seems to be easier to be deployed under a linux environment than for Windows... or maybe it's just me.
Thanks
No. The VM would need full control of the graphics card. (This can be done with most VM systems, but then you wouldn't be able to use the graphics card on the host)