A million fiber-to-the-home (FTTH) subscribers is quite a remarkable landmark: Two industry groups note that 6m homes are now passed by FTTH, and that subscribers increased 50 percent in the last six months. Homes passed jumped from 2.7m in late 2005 to 6.1m today. The associations estimate that US broadband penetration is just 44 percent with FTTH being one percent of that; dial-up only is 28 percent, the same as no Internet access at home at all.
Although this is a site devoted to wireless data, some readers would be surprised that I am a great supporter of FTTH. In any comparison, wire beats wireless for exclusive bandwidth delivered for the same adapter cost, usually by a speed factor of 40 in LANs. For instance, 802.11g delivers 54 Mbps of raw bandwidth; gigabit Ethernet, with a built-in adapter or a cheap add-on card deliver 1,000 Mbps of raw bandwidth.
Fiber can carry an awesome capacity and have its capacity upgraded by changing out components, not fibers, making it a potential multi-decade technology to the home as copper and coax have ultimately proven to be. The ultimate home connection option should, in fact, be fiber, because there's nothing better.
Now the reality. Wireless has a leg up in most large-scale deployments because of the tremendous cost of bringing fiber to the home. Companies like AT&T are looking at large-scale FTTN (fiber to the node) in which the fiber is brought to the neighborhood, and then a short-range wire is used to connect to the home. It's cheaper, but it restricts the ultimate bandwidth fairly tremendously. It also leaves out intracity and intra-LAN services that could work over unused bandwidth on FTTH that isn't provisioned for pure Internet or pure video services.
We're seeing lots of Wi-Fi networks get pitched as what I call best-worst networks for cities. They're remarkably cheap to install compared to fiber, even as they deliver potentially anywhere from 1/1000th to 1/50th of the FTTH or FTTN potential bandwidth. Wi-Fi is best-worst because it's the best we have (ubiquitous, no license, no ownership of the band), but worst because it's not designed for this purpose, and thus an industry has grown up around trying to make metro-scale service work without breaking compatibility with existing Wi-Fi adapters.
Most broadband technologies now, wired and wireless, are simply patches on not being able to afford installing fiber to every home even in major urban areas. Some would say that this is a conspiracy of sorts; Bruce Kushnick argues that telcos received billions of dollars to build promised networks that have never materialized and the money is strangely not being returned. But no one denies the cost and hard physical work to roll out 85m households with fiber.