In 1965, Intel co-founder Gordon E. Moore successfully predicted semiconductor processing power would double about every two years. A trend similar to Moore's Law is now occurring in fiber optic capacity. And just in time as this New York Times article notes, pointing to burgeoning demand for Internet bandwidth:
The need for core network improvement is pressing, said Stojan Radic, a professor of electrical engineering at the University of California, San Diego. “We are looking at a point soon where we cannot satisfy demand,” he said. “And if we don’t, it will be like going over a cliff.”
Demand is continually growing, somewhere below street level, as details of our e-mail, bank balances and national security zip along on light waves. And consumers can’t get enough video clips on YouTube, television shows on Hulu, and movies streamed to them by Netflix that they watch on their computers and TVs.
This has implications for telecommunications services, which in theory could deliver a better Internet experience and new applications with far more bandwidth. While technological advances will allow more bandwidth to move along the fiber of the Internet backbone and middle mile distribution networks, this increased capacity hits a major bottleneck at the so-called last mile that connects to customer premises.
This segment of the network is still largely made up of metal wire designed for the single purpose of delivering analog phone service or cable TV. The business models of the telcos and cablecos don't allow them to make the capital expenditures necessary to upgrade the last mile to fiber, creating an urgent need for alternative providers that can devise viable business models that can make the fiber connections for consumers.