Template Tools
| Russia passing U.S. in "fiber" |
| Written by Dave Burstein |
| Thursday, 16 February 2012 22:27 |
 All in the name. Korea’s #1 in penetration if 100 megabits on copper from fiber to the basement is “fiber.” Japan is #1 if only fiber all the way to the apartment is considered. The U.S. is far behind in either case, with the larger countries of Europe - except Russia - even further behind. Nearly 60% of Korean homes subscribe to one or the other and over 40% of Japanese. So do 27% of Lithuanians. Yes, Lithuania leads Europe. They, the Russians and other Eastern Europeans generally deliver broadband by fiber to the basement and copper to the apartment, Speeds are often 100 megabits; Russia often is near the top in average Internet speeds. "Fiber" in the U.S. only reaches 8% of homes. The vast majority of U.S. fiber lines are Verizon, which has essentially stopped building. Russia is at the same level, expanding rapidly. China is only at 4%, although they are expanding at a rate of 10M a quarter. The charts below, from iDate and the Fiber to the Home Council, are worth a look. Analysts have almost all agreed that "fiber" described a connection that was fiber to the apartment or building. That could easily deliver 50-100 megabits, even if the last few hundred feet were copper. Cable until widespread DOCSIS 3.0 topped out at 10-20 megabits.
Nearly all of us agreed that AT&T's brilliant relabeling of DSL from a field terminal was "fiber to the node" was fundamentally misleading. The vast majority of AT&T customers had service at 6-10 megabits down, much slower than fiber. The network was built to a 25 megabit standard, but all but the first 10 megabits was generally used for AT&T video. Customers couldn't access most of that spectrum. On the other hand, few customers are demanding more than 10 megabits so far. With DOCSIS 3.0 now proven to deliver 50-100 meg fairly reliably, analysis and policy becomes more difficult. Similarly, as vectored DSL becomes common at 40-80 megabits in several European countries, the line becomes harder to draw. Meanwhile, politicians in Europe and the U.S. are promising 100 megabits without any details of the real speeds to deliver. Is fiber necessary?
I still think fibering the world a good idea in the long run, but there’s no denying both DSL and cable are getting darn good. Graham Finnie points out Europe can reach 2020 goals drawing heavily on “cable broadband and DSL lines with vectoring.” http://bit.ly/xM9F7j Europe intends to bring 100 Mbit/s to 50%. DOCSIS 3.0 does that fairly well by bonding four 8 MHz channels to share 200 megabits and bonding eight channels is already built into some gear. Since about half of Britain and Germany has DOCSIS already, only a limited number of lines will need upgrading. Vectored DSL will bring 100 meg only to those very close, but can easily deliver 30 meg - the EU goal for everyone. Fiber has far more headroom. Today’s fiber is routinely a gig down and far faster than anything else on the upstream.   |