ON TODAY'S BROADBAND LIMITATIONS

WALK BEFORE WE RUN...

A few days ago, my wife's laptop drive crashed while on a business trip to London.  Having secured another PC, she called asking me to send her backup files as quickly as possible for some meetings in the morning.  No problem, I said, with my state of the art PC, and premium broadband service from Verizon, I should be able to send the 8 gigabyte set of files in short order.   

Five hours later, after having explored and subscribed to a range of third party services, that cumulatively cost over $200, I was able to send the necessary files to her.  She then took another three hours downloading the zipped files over her hotel "broadband" connection.  Essentially an all-nighter for her and an all-evening for me.   (The picture is of a British Telecom "Broadband" phone booth by Davido from Flickr...click the link for a larger image).

The experience hammered home a critical fact about today's broadband...although we're so better off from the days of dial-up, and can download music off the internet pretty easily, we've got a long way to go in sending digital data to other people; the bottlenecks include:

1.  upload limits put on by broadband providers,
2.  attachment limits put on by email providers,
3.  storage limits and charges by online hosting services,
4.  the limits of some of the prevalently used software, and
5.  the latencies driven by multi-node nature of the Internet itself. 

If you're curious about your broadband speed, both up and down, test it here at DSLreports. 

So, the broadband picture is murkier than it looks on the surface.  First, let's look at the broader broadband surface.

From Tom's Hardware, here is a good snapshot of the current status of wired broadband deployment in the US and abroad.  We've come a long way from dial-up over the last decade, crawling before we could walk.

Not many surprises overall in the report, but some observations that may be interest:

1.  Although the US is not in the top ten broadband countries, it still leads in absolute number of lines, although China is a rapidly closing second.
2.  As expected, South Korea and Japan rank up there in both density and absolute numbers.
3.  Europe is still lagging in relative terms.
4.  Not surprisingly, due to relative strength of telco monopolies/oligopolies and the relative under-penetration by cable overseas, DSL leads cable modems in most markets.

The report is excerpted below for those interested in the specific data:

Broadband continues its rapid growth: 26.5 million new lines were installed in the second have of 2004 alone, bringing the total number to 150.5 million. By the end of this year, there will be more than 215 million active broadband accounts, market researchers said.

Point-Topic, a British market research firm that tracks the worldwide development of broadband connections, reported high-speed Internet connections to gain popularity at an increasing speed. For 2004, more than 50 million new lines were installed on a worldwide basis, breaking the 150-million mark by the end of last year. Between July and December, 26.5 million new accounts were activated, representing the biggest half-year worldwide increase to date, according to Point-Topic.

The United States remain the world's largest broadband country with 33.9 million lines. China and Japan however are closing. China added more than 6.6 million new lines in the second half of 2004, bringing the total to 25.8 million. Japan reported 18.1 million lines and South Korea 11.9 million. Germany leads total broadband installations in Europe with 6.9 million lines, but showed slower growth than France and the UK, which now have 6.8 million and 6.1 million lines, respectively.

DSL has a majority 62 percent of world broadband market share compared to a 38 percent share made up of cable modem and other technologies. DSL outpaced cable also in US with 20.1 percent compared to 13.8 percent growth in installations in the second half of 2004. However, cable modems are still the dominant broadband type in the US with 20.2 million lines (DSL: 13.7 million).

In terms of market penetration, the US did not make it into the top-10 broadband nations. South Korea leads the ranking with 24.8 broadband lines per 100 people, followed by Hong Kong with 21.9 and the Netherlands with 19.

Point-Topic analysts said they expect growth of broadband to decline in percentage terms, but increase in the absolute number of lines added for the foreseeable time. By the end of this year, total broadband installations will have reached 215 million (140 million DSL lines), according to Point-Topic.

Although these types of reports are helpful, it'd be more useful to see additional, practical data points measured in future surveys.  Specifically:

1.  What are the average speeds (downstream and up) prevalent in each country at the average mass market pricing?

2.  How do the various countries and regions rank on the average speed metric?

3.  What is the ranking on the upstream metric? This can indicate higher prevalence of consumer driven content traffic.

4.  What is the ranking globally on monthly price of broadband?

Second, although the world has come a long way from dial-up, we still have a long way to go on broadband deployment.  Today's speeds already feel like dial-up speeds did a decade ago, when it comes to uploading and sharing large files.   Xdrive, a provider of online storage and backup services, has a good explanation on the limits of uploading in today's broadband environment.  This then makes you feel a little bit better rationalizing how much you'll have to pay for their services...to be fair, their prices though high, are representative of the industry at the current time.  These costs for moving your files around are likely to come down once you apply Google's server farm economics, and paid-search based revenues to the equation, much as they have for email storage.  This is even more imminent since Google hinted at providing personal video services a few days ago...and competitors will likely follow.

Increasingly, these files have less to do with illegal file-sharing and more to do with legitimate, consumer originated content like photos, movies, general data,  and "fair use" music sharing.  A lot of data traffic will be generated just to keep back ups and for convenience purposes.  Examples of this would include consumers wanting to back up their music collections in multiple locations, consume that music on various types of cell phones, PDAs, and other devices in different locations.  Other examples include sharing photos and videos with friends and family around the world.

These files are increasingly taking on multi-gigabyte proportions.  Even at a time, when web-based email services like Google and Yahoo! are providing up to 2 gigabytes of storage free, it's increasingly taking hours just to upload that much data to the servers due to upstream restrictions on today's broadband.  And it's not just about how much broadband you've got...most Internet (Yahoo!, Google, MSN, AOL and the like) and corporate email systems limit attachments to 10 megabytes or less.  You have to revert to new, third party services like Streamload to send bigger attachments and pay extra to store bigger amounts of data online for downloads by friends and families.  This consumer machine to machine (M2M) traffic is going to be different than enterprise and governmental M2M traffic; it will have high redundancy, will often be less efficiently routed, and will be a constant challenge to cost and price accurately given the myriad carriers, portals, service, device, storage, software, and content providers involved.

Download speeds also have to increase substantially to achieve the true promise of video over the Internet.  And of course, we've said nothing here about wireless broadband, which has it own sets of issues and opportunities. 

Finally, wireless could be an even better solution longer term under a changed regulatory environment of open spectrum, explained eloquently here.  Potentially, software-defined radios(SDRs)  have meaningful cost and performance advantages vs. today's electronic radios for consumer and commercial applications. 

But making regulatory changes, especially curtailing or dissolving federal programs and entities like the FCC are more likely to happen when hell freezes over, or more likely, when technology and consumer forces make it massively, politically  obvious that the need for spectrum allocation has come and gone.   

Until then, wired and/or wireless, it'll be another long decade trudging to broadband nirvana...

Comments

[Sunil Chhaya]

I'm going to have a portable multi-hundred-gig backup drive that I intend carrying with me that has the data of my hard drive backed up through IEEE1394 port every night, religiously. Public-domain broadband, IMHO, is way too slow and unreliable, for any mission-critical work. The standard sized ones are about a buck a GB (I have a Maxtor 300GB at home that I use to back up a couple of my notebooks) and the smaller, sexier ones from Lacie, for example, may be about 2 bucks/GB, and all have a USB 2.0 interface w/ most also having the firewire capability. I particularly like Maxtor's 1-touch simplicity w/ backing up data once you have installed and configured the companion (provided) Retrospect utility.

[Michael Parekh]

Good suggestion...I have the Maxtor as well...the solution above solves the backup problem, but not the sharing one...and as such the public domain internet will still need to scale to address the sharing problem.

[T1 Connection]

Excellent comments about today's broadband.
T1 Providers provides the best and fastest of T1 internet broadband.