Eugen Leitl wrote:
My (minor) beef with it is that while you offload most of heavy lifting to photonics you still use electronics and lookup.
Because for non linear operations, electronics is a lot better than so linear photonics w.r.t. speed, power, size etc. And, it's not my idea. See 'The "Staggering Switch": An Electronically Controlled Optical Packet Switch' by Zygmunt Haas, which mentioned "almost all optical" in 1993.
It is however reasonably easy to do everything at effectively L2 with a photonic crossbar if you encode geography in the headers (you have a direct proximity metric on your link slots).
How can you say BGP, then?
(You can actually prototype this with Ethernet MACs, as 2^48 in square meters happens to be half the surface area of the Earth http://www.wolframalpha.com/input/?i=2^48+m^2 So MAC collisions are not very probable, if distributed optimally ;)
The problem with large number (beyond size of CAM with reasonable power consumption) of MAC is that hash table is necessary, which means route look up time is not bounded, which means fiber delay lines can not be used. Thousands of MAC addresses in a small L2 WAN is fine, except that BGP does not work.
If you do it in optics the protocol is completely different from IPv4/IPv6,
What I have shown is that what will be completely different will be L2. IPv4 uber alles. Masataka Ohta