Broadband Speed and Moore’s Law: A Response to Robb Topolski


Robb Topolski recently penned a critique of metered pricing in which he takes issue with Internet connection speeds in an argument based on Moore’s Law. His thesis:

Moore’s Law, and its corollaries, all indicate that technology grows cheaper, or its capability increases, by a factor of two every 24 months or so. Networking technology is no different. Has your broadband bill gone down by half? Is your ISP bringing you about twice the speed and capacity than it brought to you two years ago? If not, then its a good bet that your ISP’s costs of delivering the same level of service to you have dropped during this time and they simply would rather not increase the network’s capacity as fast as you would like them to.

Moore’s Law, by way of background, has nothing to do with broadband connection speeds. It isn’t even a "law" as much as a collection of articles dating back to 1965 that taken together form "the law”.  The closest thing to a concrete statement of the principle comes from an April 1965 Electronics Magazine article:

For simple circuits, the cost per component is nearly inversely proportional to the number of components, the result of the equivalent piece of semiconductor in the equivalent package containing more components. But as components are added, decreased yields more than compensate for the increased complexity, tending to raise the cost per component. Thus, there is a minimum cost at any given time in the evolution of the technology. … The complexity for minimum component costs has increased at a rate of roughly a factor of two per year (see graph on next page). Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years. That means by 1975, the number of components per integrated circuit for minimum cost will be 65,000.

In 1975, Moore changed his theory from one year to two years. However, he never attempted to apply it to networks. While some technologists have since attempted to apply "Moore’s Law" to just about everything (the "corollaries" Topolski mentions), there is no accepted "law" that dictates the doubling of broadband speeds every two years.

In other words, Topolski is attempting to make up a new theory to justify his policy positions.

However, even if Moore’s Law did apply to Internet connection speed, Topolski would still be wrong. He falls victim to the same arguments that cripple anti-global warming arguments – it takes one data point in a series and attempts to extrapolate it to a larger trend.

You can’t say global warming doesn’t exist just because it’s particularly cool today. It just doesn’t pass the smell test. Topolski’s theory, however, attempts to do just that.

If you look at the larger history of Internet connection speeds – rather than one two-year period – you’ll see that they have steadily increased. Rather than looking at one data point as Toploski suggests, you should view Internet connection speeds over time.

Verizon illustrated this point with a recent post on their site that looked at the history of capacity and speed development in internet connections. I have borrowed for the table below, their starting point of 300 baud modems in 1979. I think that is a reasonable point at which we can begin our discussion.

Using Topolski’s “Law” of Doubled Internet Connection Speeds, I will fill in, starting in 1979, what the "doubled" speeds would be equal to, and the commercially available speeds (either based on early modem speeds or offered broadband speed) over that same period of time.

Baud Rate
Modem Equivalent
} 1200 bps
} 2400 bps
} 9600 bps } 14.4 kbps
} 28.8 kbps
} 56k
} 384k DSL
} 768k DSL
} 1.5 Mbps DSL/Cable
} 3-5Mbps cable
} 5-15 Mbps Cable/20 Mbps fiber
} Cable Wideband

If you look at the increase in capacity at the consumer level over the last 29 years, you’ll see that Topolski’s argument is in error. In fact, Internet connection speeds in the US have kept pace with the upward projection over the past three decades.

Will that trend continue? That’s difficult to say. If you assume that Internet connection speed as a broad category is the appropriate metric to use (as opposed to a particular type of connection) it’s likely that advances in technology will continue. Whether a particular technology scales up is a different question. The telephone companies found that they could increase the capacity of DSL lines only to a certain point due to the capacity of twisted pair copper wire. They began fiber deployments to increase their speeds and are currently offering 20Mbps connections where they have upgraded their capability.

Cable companies have begun the rollout of wideband modems with speeds up to 50 Mbps (a speed Topolski’s imaginary law dictates should not be achieved until 2014-2016.) Further, the channel bonding technology that enables those speeds is not limited to 50Mbps. Channel bonding can provide speeds well in excess of 50 Mbps because each bound channel is the equivalent of roughly 40 Mbps. The speed limit announcement at last years Cable Show was 160Mbps.

Under Topolski’s "Law", we should not achieve 160 Mbps until 2017-2018.

Does everyone have access to a 20 Mbps or 50 Mbps connection today? No. But Moore’s Law also never claimed that the moment a capacity doubled it would be instantly available to everyone.

As an example, not everyone had a 56k modem the moment they were available. In fact, at the time 56k modems were rolled out, the capacity had increased, but a fight over the modem standard that should be adopted delayed widespread adoption for some time. In other words, real world business decisions impacted the adoption despite the availability. Early adopters paid a premium and settled for modems that were often incompatible with their ISP until the v.90 standard was announced.

Similarly, there is not a single cable infrastructure. Different companies run on different networks and prioritize investments in those networks based on usage and cost. Business decisions still drive availability and adoption regardless of what may be technologically possible. Are most cable customers able to access the Internet at 5 Mbps (the Topolski’s "Law" suggestion for where we should be today). For the most part, yes. Again, though, different companies invest in upgrades to their network at different rates.

The fact is Internet connection speeds are increasing, prices are generally flat or declining, and all of this is being done despite what Topolski’s improper application of Moore’s Law would suggest should be the case.

  • http://CableTechTalk Ken

    “Back in the day” (less than 10 years ago), the best I could do was $22 for 56Kb dialup, plus $15 for a 2nd phone line, for a total of $37 a month. Now, I pay $42 a month for 6Mb access and $3 for a leased modem. So, for $8 a month more than I paid 10 years ago, I get 100 times the speed, and the price of broadband service hasn’t increased since I signed up. You need to put all this talk about “the ISPs need to increase capacity” in perspective. Factor in inflation, and I’m probably paying LESS than I was for dialup plus a 2nd phone line. And the price per bit is poised to drop precipitously in the next few years as more capacity comes online. Sure I want more speed, I find it hard to complain.

  • Pingback: Kung Fu Quip » Andy Kessler Gets It Wrong. Really Wrong.()

  • Bredbaand

    Broadband is beneficial as it reduces energy consumption that comes with long period of time spent on the internet. It also ensures safety as some emergencies can be solved with faster and easier internet access just like some opportunities can be earned.

  • Chicago Health

    Yay google is my queen assisted me to find this great internet site ! .

  • Jsparco

    Thanks for sharing! these were great, I'm excited to find the new Orlando Golf Course

  • Jsparco

    Thanks for sharing! I'm glad to see the cable companies atleast improving the technology. International School in India

  • london used cars

    Before we start, it’s worth noting that much of the cord-cutting coverage I see online seems to begin with frustration at prices (somehow never compared to the costs of other entertainment options) or by desired flexibility in purchasing options (they just want to get the one network or the one show).

  • gmat prep

    Attractive post. I just stumbled upon your blog post and wish to say that I have really enjoyed analysis your blog posts. I found the perfect place for my needs. Contains wonderful and useful messages. I have read most of them and has a lot of them.

  • The Dukan Diet Book

    They are seeing to it that things are not really going to get complicated. This is really a must if we want to take a leap to the next level.