Anybody who's ridden the plummeting spiral of bandwidth prices might be asking: Where does it end?
The answer: 2005, when the rate of falling bandwidth pricing stabilizes, bankrupt carrier assets have been weeded out, and carriers actually contemplate increasing capital spending budgets as their revenues start to rebound. This process is already well underway.
That's the conclusion of Bandwidth Price Revolution, the latest issue of Light Reading Insider, Light Reading's paid subscription research service.
Taken in a historical context, the massive bandwidth glut has something in common with all technology revolutions: The early stages are characterized by oversupplies and bubbles. The second phase, which comes after the initial glut is worked off, is characterized by a new round of innovation and growth.
The new report concludes that the second stage of the boom is likely to warm up in 2005 as the bandwidth/capital spending cycle begins its turn. At that point, carriers will have started ramping new services that take advantage of cheap bandwidth.
The most successful carriers will be those that move into Ethernet-based data services and successfully market new services tied to the bandwidth delivered by Ethernet connections. In the short term, such moves are proving painful, but they will pay off in the long run.
One of the winners may be AT&T Corp. (NYSE: T), which is taking a hit now because it's making some decisive moves into a multiservice network that can handle new data services.
Sure, AT&T's revenues have fallen for 15 straight quarters, nearly four years in a row. And, sure, the stock price has taken a beating second to none.
But AT&T has an edge in that it has stabilized profits and cash flow, while at the same time aggressively embracing metro Ethernet services. AT&T is trying to emerge as a leader in the corporate data services market. And it's trading at a modest multiple of only four times sales, making it an attractive target for somebody looking for a multiservice newtork.
The report points out the the recent mauling of core bandwidth providers shows that metro data services look to be the savior, while wholesale optical circuits clearly are not.
One brief example of a company that helped create the bandwidth glut (and did nothing to profit from it): During the Internet bubble, 360networks Inc. spent nearly $1 billion to build a network that carried 192 Gbit/s in capacity. In the end, 360networks alone could have carried all of the traffic on the Internet. After spiraling into bankruptcy, its assets were sold at a tiny fraction of their original build cost.
What about those telecom assets? They're being sorted out, more quickly than it seems. A recent shuffling of debt and assets has started to put the sector back on track. In fact, the Insider details some 18 transactions that have occurred over the last two years.
One example in the asset pick-up category: Corvis Corp. (Nasdaq: CORV), through a controlled company, picked up Broadwing's network from Cincinnati Bell in June for $91 million, a fraction of the estimated $2.1 billion spent to build the network in the first place.
Table 2: Recent Asset Transcations
Asset
Date
Acquirer
Seller
Amount of Transaction
Est. Cost of Building Asset
Broadwing Inc.
June 2003
CII Communications LLC (controlled by Corvis Corp.)
The current Light Reading Insider report – Bandwidth Price Revolution – is available here. A single-user license to the report is $400. An annual single-user subscription to the Insider, which includes access to the complete archives, the current report, and each of the monthly reports issued over the next 12 months, is available for $1,250 per year.
re: Report: Bandwidth Glut to End in 2005 The LightReading note on the report indicates that new services are going to use up the existing glut and drive demand for even more bandwidth. However it seems to neglect indicating exactly what these new services will be. It would be very instructive to see a list of the new services that IP technology will drive. I have been working in this field for quite a while and the services that I see as being enabled do not drive much bandwidth utilization. Other parameters such as connectivity and latency are much more important and do not put an excessive demand for bandwidth on the network.
I agree with the note that there will be a wave of innovation to take advantage of the opportunities being produced by the new technologies. However I do not see much bandwidth being consumed in the core of the network by any of these services. The glut will of course depress bandwidth prices and shape how services are designed. However there is no crying demand for bandwidth from any of the useful new services that are on the horizon.
Recall that the most successful existing service is Email. Most new services will bear a closer resemblance to Email than some of the bandwidth hogs that people talk about but never result in paying customers.
re: Report: Bandwidth Glut to End in 2005 I have not read the LRI report, but from the above summary it seems to me that health of the sector is equated to a resumption of spending on backbone capacity.
I would like to contribute the following observations:
1. Demand for backbone capacity is less than linerarly proportional to the number of users times their access bandwidth. This direct proportional relationship between capacity and number of users decreases rapidly as you go further into the network, as a result of overbooking factors, the law of averages and bandwidth-saving devices such as caches and even store-and-forward mail servers (which also help to spread the load in time).
For some of the above reasons I would also expect demand for backbone capacity to continue to drop as a percentage of access capacity.
2. Backbone operators who have invested in infrastructure during the boom months are and have been suffering from massive price erosion in at least to ways: overcapacity has caused prices to tend to the incremental cost of providing the service (or even below this cost), while at the same time vendors have continued to innovate which has continued to drive down the above-mentioned incremental cost.
My conclusion is therefore that even when at some point in the future demand for backbone capacity will start to outstrip installed capacity, the cost per bit of new capacity will be so low that there may be a danger of a renewed glut as backbone operators (who still own masses of dark fibre) once again try to increase their market share by simply making capacity available. Moreover, the need for backbone capacity as a percentage of users' total bandwidth needs will continue to fall.
The future of our industry is therefore to be found in access and in the edge, not in the core.
re: Report: Bandwidth Glut to End in 2005 Author: mrblobby Number: 2 Subject: Relevance of backbone spending as industry health check Date: 10/29/2003 7:54:23 AM --- The future of our industry is therefore to be found in access and in the edge, not in the core. ---
AMEN, brother. As a large global enterprise customer, I'm seeing very little cost reduction globally even though it may have plunged amongst the Telecomms and in the backbone. Access continues to resist price declines and is my core barrier to adding bandwidth to my infrastructure.
And in PTT monopoly and developing countries, there is really nothing any of us can do to spark changes in access...
re: Report: Bandwidth Glut to End in 2005 I don't think LR is making the case that new bandwidth-intensive services will chew up all available bandwidth. The case seems rather that as bandwidth prices stablize, it will be possible for service providers to finally offer services profitably.
Siemens has a chart they're fond of showing that points out the revenue per MB of various services: Internet Access $0.02 USD Local calling $0.04 Long Distance $0.12 (European rates I presume) SMS service $781.00
Thats an interesting reminder that its not necessarily bandwidth that will generate revenue but services that are useful to people; mobility, ubiquity, presence, etc. Other services will depend on QoS, latency, jitter, etc and they will command premium prices. Its a matter of building a network that will deliver all this instead of creating a huge bandwidth-intensive network.
re: Report: Bandwidth Glut to End in 2005 dljvjbsl said: "the most successful existing service is e-mail."
Huh? In my experience, web content delivery accounts for at least 100 times as much bandwidth as e-mail. I don't know of ahjybody that shifts from modem to DSL or cable because of e-mail.
re: Report: Bandwidth Glut to End in 2005 Is this the dark fiber glut, the lit fiber glut, or the long-haul IP/MPLS capacity glut?
My guess is that in the long haul, including transatlantic and transpacific, there will will be no need for additional dark fiber for the foreseeable future (say until 2010) except for a few new routes. If you own a piece of installed dark fiber, don't expect someone to pay you more for it in 2005 than you can get today, which is pennies on the dollar.
For lit capacity in the long haul, my guess is that the price/Gbps you can charge will continue to decrease at dizzy pace, but the overcapacity will continue to make this a zero-margin business for the forseeable future as technology advances continue to drive down the cost of lighting 10Gbps circuits. 10Gbps will remain the sweet spot for the foreseeable future.
For IP capacity, we don't have as much of current glut. The core ISPs spent all their money on dark fiber and lighting the fiber, so they stretched their router CAPEX cycle and are now running their routers at higher capacities than they were historically comfortable with. Unfortuantely, due to last-mile constraints, traffic growth is now slower than Moore's-law cost/performance improvements. This means that the overall revenue per year will decrease in the core IP space.
re: Report: Bandwidth Glut to End in 2005 Huh? In my experience, web content delivery accounts for at least 100 times as much bandwidth as e-mail. I don't know of ahjybody that shifts from modem to DSL or cable because of e-mail.
The success of Email is revealed by the universality of its use. Its success is determined by the utility that users find in it. Everyone uses Email therefore Email is very successful. Indeed Email is the most succesful application on the Internet.
Bandwidth usage is not a definitive indicator of the success of an application. Most new applciaioisna re not heavy users of bandwidth. The bandwidth used by a VoIP service per user is very low. It is connectivity and latency that counts as indicated by the Email example.My opinion is that the bubble burst because the industry did not realize this.
re: Report: Bandwidth Glut to End in 2005 arch_dude
I was just thinking aloud :-)
Knowing the price pressure the market is facing, and due to the fact that we don't who these 2 suppliers are, I was wondering that they must have proposed (and will charge) an arm and a leg for these fiber plants since this is a govt. contract.
Hence these two suppliers revenues will be higher than their competition if we compare links for links in the same geographical areas.
I may have gone on a tangent but it is something to ponder..........
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
(17) | 
Digg
Del.icio.us
Reddit
Tweet This