Improbable as it may seem, but the bandwidth glut created by the telecom bubble of the late nineties might be coming to an end. Or there is at-least light at the end of the tunnel. Many of the major long haul service providers are in fact are in the process of lighting up their dark fiber and/or additional wavelengths to add capacity. There is talk of a new transpacific cable being laid by a new group. The network operators who are looking to add more capacity include VSNL, FLAG Telecom, Asia Netcom, and Telefonica.
But unlike the 1990s, this time around the capacity is being added in a as-needed basis, according to research firm, Telegeography. The demand is being driven by a big growth in consumer broadband connections, says Alan Mauldin research analyst with Telegeography. The growth has particularly strong in Asia, and Latin America, where traffic has more than doubled, Mauldin says.
Point Topic, another research firm estimates that there were 209.3 million global broadband users at end of 2005 up 56.2 million or 37% from 153.3 million lines on 31 December 2004. In some countries, such as Brazil, Russia, India and China, the demand for high speed connections is still in early stages, and that is why many in the industry are hopeful that the glut might be over.
The best news however, is the likelihood of a balance between supply and bandwidth demand. A lot also has to do with rapid consolidation in the telecom industry. AT&T+SBC+ BellSouth, Verizon+MCI, Tata Telecom+VSNL+Teleglobe+Tyco Network, and Reliance-FLAG Telecom are some of the key deals that have managed to take off excess capacity off the market.
Lately, Level 3 has been on a buying binge and has acquired Wil-Tel, Progress Telecom and last week there was word that they might acquire ICG Communications. (The news could be announced this week, some say.)
In addition to the consolidation, the demand for bandwidth for new fangled applications like IPTV, VoIP, build-out of new networks are helping equipment providers turn the corner, but the wholesale bandwidth prices are still too low for big backbone companies to turn a profit. And when that happens, perhaps we can be comfortable in publicly admitting that the telecom turn around is in full swing.
If The New York Mets can be 9-2 at the end of first two weeks of the new baseball season, then anything is possible.
13 thoughts on “Bandwidth Glut Almost Over?”
Does this mean that fiber will return to Moores Law as new multiplexing cranks up capacity per fiber strand and repeaterless hops drop the cost of long distance cable?
Another factor is that in spite of the whining of the carriers, voip/cellular and time shifted video are (if my memory is correct) much more bandwidth-efficient than circuit switched voice and “broadcast” video thus the shift to IP creates some of its own capacity. The real problem is of course migration to IP.
I suppose the good-news story here is that less competition will allow “managed” deployment of technology (remember the first digital COs?) and coupled with “increased margins” will lead to more gradual price declines and a more stable industry.
Nah! The phone/cable companies haven’t changed a bit and the VCs will “discover” Fibr 2.0 and the capital will start lining up for another bubble … er boom.
Works for me.
this was just a view of the trans-oceanic links – what about the massive overbuild of fiber running between cities and coast to coast here in the US alone? seeing a bite taken out of that excess capacity would REALLY impress me.
from the looks of it, it seems that the US capacity is being worked out as well. the fiber providers are collapsing quite rapidly – the VZ, MCI etc etc is going to have an impact.
in fact the metro networks are getting full as well, and i have reported that in the past.
profits, however remain quite elusive.
I see demand for bandwidth continuing to grow with the increase in consumer applications like IPtv. It seems that the capacity is there and now has the potential to be managed efficiently. That makes me even more frustrated with the franchising bottleneck that is holding up consumer access to bandwidth heavy offerings.
Interestingly enough, the file name of the price graph in this piece is:
Having said that, ISPs in New Zealand tell me that international transit pricing is going up for the first time ever. This is apparently because MCI-Verizon has decided to “dump bandwidth” on the Southern Cross Cable, of which it owns 10 per cent.
Many carriers are lighting additional wavelengths on existing long-haul fiber to meet demand. Thats different from lighting currently dark fiber. And its very different from needing more fiber, except, perhaps, on some routes that never had enough capacity like trans-pacific.
It does cost some money to light up waves. Not much, compared to the cost of trenching new fiber, which is extreme. Even when we need more fiber, there is a massive amount of empty conduit space through which fiber can be cheaply blown.
Finally, there have been no major advances in wave division multiplexing technology in 5 years because of a lack of demand for greater wavelength density. When we need more capacity on fiber, the engineers at Lucent, Nortel, Ciena, et al will pull out their plans for higher capacity DWDM gear and start selling it.
Yes this is all true.
I’ve even had conversations with some RRs (Norfolk Southern and CSX) who’ve run out of dark fiber on some links, and (gasp!!!) have had to rent capacity from the RBOCs.
Is Fiber just like the Oil business? ie, it goes from a top to bottom every 5-10 years. Typical game theory as to who makes the big CAPEX bets.
Now the question is, will these guys get their act together and slow investment to keep the prices high?
I mean can the government force a business to make CAPEX purchases?
thanks for pointing out the lighting up additional wavelengths. for some odd reason i had blanked out on this and well, please accept my apologies. hopefully the correction will help clarify.
Ok. Reality check. How much do you think it would cost AT&T to run a new fiber optic cable along the floor of the PACIFIC OCEAN?
Now, how much could they go spend on, oh, nothing fancy, say a DWDM multiplexer from Black Box, and put it at both ends of an existing fiber. Now they have 16/32/64 wavelengths, or 16/32/64x the capacity out of that fiber that is already laid. Now suppose they don’t go with Black Box, but go with someone a little more reputable, like Huewei or Ciena?
Play SimCEO with your SimTelco with that scenario and see if this article isn’t just FUD. Saying certain Telcos want to add existing capacity is one thing, but saying they want to lay new fibers across the ocean is more of a stock tip than a technical article.
reading all these comments tells me NOBODY really has a grasp of what truly is going on with “”LIT”” vs Dark Fiber and everyone just ASSUMES you can light up anything you want…capacity of some routes ALREADY MAXED OUT and full…this means CAP EX to light up more lines and nobody has CASH
Has anyone been trenching empty conduits in the last 6-8 years. I don’t believe so. Back in Nov. 2000, Jim Crowe, CEO of Level 3 wrote:
“Today, we believe that there are only twelve empty nationwide conduits and Level 3 has eleven of them.”
Looks like the “massive amounts” of empty conduit space that Dan talks about, all belong to the same company. The other empty conduits are at best a patchwork quilt.
Seeing a bite taken out of that what capacity? What excess capacity? For the US to truly grow technologically, all those oc-48s and oc-192s will be necessary if the rbocs would actually increase the capacity of the individual subscriber line. They always give more download, never upload. This is because of the limited capacity to the major junctions. If Viop, Video and application hosting are all going to be more intrinsic to data services, then all those old technology lines and equipment go away. bell has made billions off of the subscriber lines with very little move forward. Thats why wireless is moving in leaps and bounds…its the last mile, not the back bone. Turn up the last mile and all that dark fiber will need added capacity.