Kepler is based on 28-nanometer (nm) process technology and succeeds the 40-nm NVIDIA Fermi architecture, which was first introduced into the market in March 2010.

ENGADGET: Primed: why nanometers matter (and why they often don't)

In other words, they could adjust the beam to only pull the 400 nm particles, then change it to only pull the 1000 nm particles.

FORBES: Scientists Build A Working Tractor Beam

Scaling the capacity with the line width indicates that 14 nm chips could have capacities of about 774 Gbits and 10 nm chips could have capacities of 243 Gbit.

FORBES: Rare Gases and Fine Lines