Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

Is it impossible to pass and manipulate a multi-dimensional array in CUDA, or is it intentionally avoided for the sake of performance?