The current generation of switches that we're working on uses diffraction gratings patterned onto glass hard drive platters, installed in a modified hard drive, spun by a custom motor controller that's synchronized to the NICs via PTP.
Ha. I was going to guess it would be an AOM, I wouldn't have guessed a diffraction grating on a hard drive platter. That's awesome, and must be incredibly energy efficient.
The cost of switch ports and interconnects could all be dumped into making host interfaces faster, allowing for the switching time to be reduced.
Crosstalk is better than 30 dB, and double pass loss between ports is 5-8 dB. The switch is basically cycling through three or four different interconnection patterns that are defined by looped back fiber connections, so the signal has to pass through the switch twice.
The cost of switch ports and interconnects could all be dumped into making host interfaces faster, allowing for the switching time to be reduced.