---------- Forwarded message ---------- Date: Tue, 17 Mar 1998 16:49:52 -0500 (EST) From: [EMAIL PROTECTED] Subject: One cent per minute IP telephony For more information on this item please visit the CANARIE CA*net II Next Generation Internet web site at http://www.canarie.ca ------------------------------------------- I, CRINGELY "The Coming World of One Cent Per Minute Long Distance: Why Traditional Phone Companies are Scared of the Future Even Though They Should Not Be" Computer industry gadfly Bob Cringely looks at how Moore's Law will apply to the communications industry in the next few years. http://www.pbs.org/cringely/ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ The economics of computing are different from that of almost every other industry because the price of computers is nearly always going down, not up. The first person really to quantify this was Gordon Moore, co-founder of Intel. Moore noticed that the number of transistors that could be packed on any given area of silicon in an integrated circuit seemed to double every 18 months due to technical advances in manufacturing. What really cost money was not the number of transistors, but the area of the silicon wafer, which generally remained about the same cost. The number of transistors, which represented the power of the circuit, doubled every 18 months -- meaning that users could get twice as much performance for the same number of dollars per acre of silicon. Looking at it another way, a smaller area of silicon could now support as many transistors as before, allowing the total circuit to be smaller and cheaper. That's Moore's Law: We either get more power for the same money, or the same power for less money, and it's been this way for well over 30 years with no end in sight. For makers of computers, the effect is that their inventory is always going down in value. In that way, personal computers are like ice: Left unsold, their value eventually evaporates. Moore's Law has generally been applied to hardware, not software or services -- until now. We are about to see Moore's Law applied to communication services of all types, with the result that phone service, for example, is about to get a lot cheaper. The key to tying a service to Moore's Law is how much that service is reliant on underlying chip technology. The more you can replace service workers with chips, the more closely-tied that service will be to the underlying value of the chips used to make it happen. Eliminate people entirely, and even the most complicated service begins to look more like a product. That's what generally seems to be happening with sending bits of data over a telephone line. For the most part, the cost of telephone service has been dropping for years. Corrected for inflation, today's $0.35 call from a payphone is cheaper than the nickel local call made in 1958. Local phone service has generally cost the same as long as people can remember, which means that it, too, has gone down in real cost corrected for inflation. And the price of long distance service has dropped in absolute terms following the breakup of the Bell System in the early 1980s. What we pay for long distance calls today is significantly less that what we paid 10 years ago. But for all this general softening of prices, the downward slope of the price line has been shallow compared to Moore's Law -- until now. Suddenly, we are in the world of dime-per-minute long distance. America Online has lately been pounding me over the head with offers of nine cent per minute long distance. Well, it is only going to get better (or worse, depending on whether you are buying or selling). If the average cost of long distance is now 10 cents per minute, and Moore's Law is finally coming into effect, then by the end of the decade, a minute of long distance should cost five cents. By 2002 the price should be down to 2.5 cents, and by 2004 it should be getting close to one cent per minute. Sounds right to me. It's not just long distance that will go down in price, but all communication services. Long distance is just an easy example, but the same thing will happen to your Internet service costs despite the momentary upblip from AOL. We can trace this precipitous price drop to two things -- a drop in the real cost of providing such services, and to increased competition. New types of optical fiber and different ways of encoding signals on that fiber are resulting in an explosion of backbone bandwidth. It's possible now to send many terrabits-per-second of data over a single optical fiber. Though the switches can't yet keep up, it is very possible to send traffic equal to every simultaneous phone call in America over a single optical fiber. And since we have lots more than one fiber connecting every city, the result is bandwidth to burn, though not all the way into our homes -- not yet. Volume buyers of data services are seeing the effect first. Between the time I ordered my DSL Internet connection and its installation, the actual cost of the service dropped by half. T-1 connections, which for regulatory reasons have generally stayed higher in price, are being supplanted by cheaper DSL lines. For buyers, the long term impact of all this is on their short term thinking. It just doesn't pay to make any long commitments for communication services in an environment where the price is dropping so quickly. While new types of fiber and wave division multiplexing are really driving down the cost of communication services, what's getting the most attention is voice-over-IP -- using Internet technology for the telephone system. New companies like Quest and Williams have been built apparently with the sole purpose of offering IP voice circuits that cost less than a tenth of earlier technology. What's odd about this is that IP telephony is not necessarily good telephony from the user's experience. The quality of an IP telephone call just isn't that great. And for reasons that have more to do with the laws of physics than those of economics, voice-over-IP isn't going to get all that much better. So these new phone companies, while they can easily drive down the cost of phone service, aren't doing much about improving the quality of that service. But that certainly didn't stop Wal-Mart, and it won't stop these IP phone efforts, either, because for some consumers price means everything. But it's my suspicion that these voice-over-IP companies are going to make most of their money in the data business, not voice. So what does this mean if you are an investor or a traditional phone company? On the surface, it looks like bad for AT&T, the Regional Bell Operating Companies, and for any other well-established carriers with older technology. But I am not at all sure I'd count out these players for two very important reasons. First, never underestimate the power of inertia. The new voice technologies will be rolled-out slower than anyone expects because consumers don't move too quickly and local phone companies have no incentive to move quickly. Second, these older companies are scared most of market saturation. Once we all have a phone line and a fax line and a data line and a cellular phone, won't we stop buying new telephone services? Worse still, these companies can't think of a way to get us to spend more time on the phone than we already are, so their revenue models show inevitable declines. One RBOC I know has looked 10 years ahead and sees the cost of long distance service at a tenth of a cent per minute. But I know that both nature and Silicon Valley abhor a vacuum, so I expect new communication services to appear that suck up that available bandwidth. I don't think it is going to be a problem at all. The phone companies, which aren't noted for being all that smart anyway, are fixated on one particular view of Moore's Law. They see their revenue for existing services dropping dangerously. But the other view of Moore's Law says that for the same money we can do a lot more work and that's where I think the market will go. People will still pay the same amount per month for communication services. And while they may not spend any more time making long distance calls, the kind of communication services they will be tending to use will require greater bandwidth. Maybe we're talking about videophones, I don't know. My strong sense is that these new bandwidth-eating applications will be in areas I haven't imagined and will make new fortunes for new players. That's just the way it has always been here in the Valley. To subscribe or unsubscribe to the CANARIE-NEWS list please send e-mail to: ------------------------------------- Bill St. Arnaud Director Network Projects CANARIE [EMAIL PROTECTED] http://www.canarie.ca/bstarn +1 613 660-3497
