To: Stash who wrote (138 ) 12/9/1998 11:03:00 PM From: ahhaha Read Replies (1) | Respond to of 626
Thanks Stan, every little bit helps. DWDM is to SR what xDSL is to cable modem. On this thread DWDM is the enemy even though it is the best technology popularly available and a vast improvement over everything else. DWDM could handle bandwidth demands for at least 3 more years before encountering load instability. At the end of the three year ramping period the cost to implement DWDM on future expected bandwidth minimums would be rising exponentially with linear increments to bandwidth additions. This level of demand will be driven by ubiquitous two n-way VOD where n-way is a reflexive topological lattice. Some will want to broadcast to everyone and everyone will want to be able to receive and send to at least one other, video simultaneously. Only SR has the potential to cope with that much bit transfer without bankrupting the entire world. INTC and CSCO are interested in chip and board level opto-electronic devices. A nanotechnological Silkroad would be an interesting development at the chip external bus level, and even more provocative at the internal level since there the thermal constraint at .01 microns for electron flow in copper is formidable, if not insurmountable. All optical devices have the potential of being shrunk all the way to the quantum mechanical regime, 1000 times smaller. That's not what INTC or CSCO have in mind though. They are only trying to interface the current DWDM state-of-the-art transmission technology with standard devices which can operate at DWDM throughput rates. Otherwise, routers and the chips in the routers become the major bottlenecks. They're already that now. SR is concentrating on the long haul pipe rather than pursue exotic on-board technology. When it comes to the network edge and the local loop, it will be some time before even DWDM gets significant penetration, but SR and other Pure optical unicolor solutions will crowd DWDM towards the last mile before sending it into technological obsolescence. Guess that doesn't say much for the prospects for SONET or, eventually, ATM.