Here's an article on how nanotubes can be grown off dendrimers:

http://www.newswise.com/articles/view/507134/

So this is beginning to look like the makings of neuronal-type
arrangements. Since nanotubes are strong, stiff, flexible, etc, why
keep them confined to a 2D-wafer surface for forming circuits?

Perhaps nanotubes are the ideal connectors for forming circuits in 3D.
Perhaps the nanotube strength could allow for them to be suspended in
a liquid or gel medium with lesser support, instead of embedded into a
solid matrix. That liquid or gel medium might then permit some kind of
transport of molecules through it to the nanotube/dendrimer network.
Hmm, perhaps eventually emulating something vaguely resembling the
neurochemistry of biological systems?

Comments?