Dark Green: Firm commitments from incumbent: BT (10M), Belgacom, Australian NBN, Swisscom, Austria, Bezeq Israel, Chunghwa Taiwan, Telus Canada, Telekom South Africa, SK Korea, (U.S.) AT&T, Century, Frontier, Windstream, Belgium, Omantel
Mid Pink: Smaller carriers in Germany, Norway, Finland, Japan
- Published: 26 August 2015 26 August 2015
2,000 homes connecting in coming weeks. Everyone in the industry is watching BT's trial of G.fast - even if BT is now deploying "G.halffast." Gavin Patterson. CEO of BT is telling investors and regulators BT will rapidly deploy 10M+ lines. In 2014, BT tested G.fast technology at 600 and even 700 megabits down on very short loops. This February, Patterson was still talking 500 megabits. Those speeds required going to many of Britain's 4,000,000 "distribution points."
This spring, they reduced the targeted speeds to 250 correction 10/10 330 megabits, going to existing street cabinets instead of the distribution points. The money men decided the higher speeds weren't worth the extra cost. That's risky - cable in England will soon be at a gigabit. Many BT customers - 5-20% - will get much slower speeds, depending on how many cabinets they ultimately deploy.
- Published: 16 August 2015 16 August 2015
230 megabits (up and down combined) looks to win at Deutsche Telekom. Call it what you like - Alcatel, Huawei, Adtran all have their own names. 35b - referring to the 35 MHz used - is emerging as most popular.
35b uses more MHz to get higher speeds out of DSL. To keep the comparison meaningful, I subtract the upstream and hence call this 150 megabits. The speed is from a KPN lab test (below, KPN pr in Google translation.) Paul Spruyt of Alcatel believes speeds still have room to improve within the 35 MHz band. Adtran tells me production units suitable for volume deployment are unlikely before summer of 2016.
- Published: 15 August 2015 15 August 2015
Gigabit required to keep up with cable. It's in the standard, but until recently everyone assumed it would be delayed for lack of customer demand. But with gigabit cable coming close, AT&T is pressing for a gigabit over DSL. Adtran, a major supplier to AT&T and Deutsche Telekom, has proven that G.fast can easily exceed a gigabit downstream by bonding two lines. It's just a lab demonstration for now without a timetable for field trials.
G.fast vectoring and pre-coding require serious processing power to keep up with the high speeds. Bonding for even higher speeds is a challenge, but Adtran CTO Kevin Schneider doesn't see any obstacles that will prevent practical deployment.
- Published: 14 July 2015 14 July 2015
BT cuts back on G.fast speeds, deployment. 150-400 down is still a big improvement over BT's current network but it will typically be half the speed originally planned. BT was going to deliver 500-800 megabit G.fast to most of 4,000,000 distribution points on light poles, in basements and in very small dug in enclosures.
Instead, they would deploy to existing cabinets (10,000's of thousands,) avoiding the cost of running fiber so many places. That saves billions but reduces typical download speed from 400-700 megabits to 150-400 megabits. In either case, 5-15% or so of homes will run even slower.
- Published: 14 July 2015 14 July 2015
Not a gigabit but pretty darn good. AT&T, assured the DirecTV merger is going through, is about to go public with the long-planned upgrade of ~5M U-Verse homes to hundreds of megabits. They will take four years and reach about 15% of their homes. Gigabit gear won't be available for several years.
Randall Stephenson in 2004 told Wall Street AT&T was using fiber, not copper, for all “new builds,” although I believe it took a few more years. “New builds” are 1-2% of the network each year. That suggests four million to eight million homes are ready with FTTB.
With most of the construction already done, the upgrade will easily fit in AT&T reduced capital budget. I estimate G.fast will cost $75-250 per home passed. The cost over four years is is only a few % of annual capex and will be recovered in months from customer charges. It would be stupid not to upgrade and AT&T's top management isn't stupid.
The 400-700 megabit down speed of today's G.fast is confirmed by the early field trials. I previously reported that Swisscom, "Gets between 285 and 402 Mbps downstream and between 85 and 109 Mbps upstream. Results depend on copper cable length: on short copper cables (ca 24 meters) we were even able to reach 624 Mbps Downstream / 151 Mbps Upstream in field but without connected customers." Telekom Austria measured 536 megabits down and 116 up.
The chart below, from Alcatel, shows combined upstream and downstream speeds in early testing. Combined, the speeds up to 150 meters are 500-800 megabits. Subtracting 100 megabits for the upstream yields 400-700 megabits downstream, what I had estimated from a different data set.
- Published: 13 July 2015 13 July 2015
G.fast uses 106 MHz compared to only 1 MHz for ADSL. More MHz means more capacity but only works over very short loops. The actual ITU standard is 300 pages with dozens (?hundreds) of features which have relatively little impact on the speed. Traditional VDSL used 17 MHz for speeds around 100 megabits. G.fast can use 106 MHz for speeds typically 400-700 down, 100 megabits up. (The "gigabit" is pr fluff today although upgrades to actually get a gigabit are in the works.)
The distance limit: High frequencies used by G.fast don't go very far in ordinary telco twisted pair cables. It was designed for 50-150 meters and speeds drop rapidly after that. British Telecom and AT&T engineers told the ITU standards group that short reach was fine and G.fast nodes would be 8-16 homes. That set the standard.
G,halffast When British Telecom realized the cost of running fiber to 4M "distribution points/curbs" they sent the entire plan back to the drawing board. The maximum wire length was increased from 250 meters to 500 meters. Chip makers are scrambling to modify their chips to the new demand. Many English homes will get speeds of 100-400 down, not close to the promised gigabit but pretty darn good.