Close

[network] The digital divide on IP addresses

(This post was updated in http://www.petersgriddle.net/2010/05/digital-divide-on-ip-addresses.html)

Every computer connecting to the Internet has to have an Internet address. No problem, you would say, unless you realize that there are more people living on the earth than there are Internet addresses.

Internet addresses are defined by the Internet Protocol (IP for short), which is currently in its version 4 (and has been since it left the labs). IP version 6 is going to solve the address space problem, if only it were deployed on a large scale. Unfortunately, it is a massive cutover, and will probably touch more computers and software than the millennium problem and the conversion to the Euro combined. And nobody seems to be in a hurry! According to Iljitsch van Beijnum, who wrote a book on IPv6: “IP addresses will never really run out, much like a tube of toothpaste, or an oilfield, is never completely empty. There is always some more that you can squeeze out.”
That may be true, but the price will go up soon. Estimates are that between the years 2010 and 2012 IPv4 addresses will become really hard to get.

So, one way of looking at the impending transition is to see where the addresses are today and who will be needing them in the near future. Iljitsch shares some thought on the latter on his weblog.

Here I want to look at where the addresses currently are. For that we took a recent look at the current (January 2006) statistics and came up with the following listing. It shows that 31 countries have more than their fair share of IP addresses, and 18 countries even have more IP addresses than inhabitants (including all babies, etc).

IP space
Inhabitants
Ratio
Vatican City
8,192
1,000
8.19
United States
1,324,925,184
278,357,000
4.76
Canada
67,430,400
31,147,000
2.16
Gibraltar
47,104
25,000
1.88
Monaco
63,488
34,000
1.87
Iceland
514,048
279,000
1.84
Liechtenstein
51,232
32,300
1.59
Finland
7,587,712
5,171,300
1.47
Australia
26,868,224
18,886,000
1.42
Sweden
12,314,752
8,861,400
1.39
Denmark
6,912,960
5,330,000
1.30
Norway
5,595,680
4,478,500
1.25
United Kingdom
73,804,824
59,623,400
1.24
New Zealand
4,436,224
3,862,000
1.15
Japan
142,997,248
126,714,000
1.13
Bermuda
67,072
65,000
1.03
Netherlands
16,400,808
15,864,000
1.03
Switzerland
6,915,008
7,160,400
0.97
Hong Kong
6,291,712
6,782,000
0.93
South Korea
41,907,456
46,844,000
0.89
Singapore
2,857,472
3,567,000
0.80
Ireland
2,942,880
3,775,100
0.78
France
45,161,856
59,225,700
0.76
Malta
283,648
380,200
0.75
Taiwan
16,280,064
22,256,000
0.73
Austria
5,788,064
8,091,800
0.72
Tuvalu
8,192
12,000
0.68
Nauru
8,192
12,000
0.68
Luxembourg
283,392
435,700
0.65
Norfolk Island
1,280
2,000
0.64
Germany
51,132,368
82,164,700
0.62
World population
3,700,000,000
6,490,757,730
0.57

 

There are a few points here: if the other countries will reach the same development level as the top 30, there won’t be enough addresses to serve them. The hunger for new addresses will be felt most severely in India (which has 1 address for every 165 inhabitants) and China (1 address per 17 inhabitants). Together these account for 2.3 billion inhabitants, in need of 2.2 billion addresses, while there are only 1.5 billion addresses left right now.

5 Comments on “[network] The digital divide on IP addresses

Anonymous
16 January 2006 at 10:29
Anonymous
25 January 2006 at 16:21

>The analysis of future need is slightly flawed though. It’s somewhat like saying at the turn of the century (the other century) that there wouldn’t be enough whale capacity to support developing countrys’ demand for blubber. By the time the problem became acute, the problem could be bypassed by going straight to the new technology (petroleum).

Developing countries don’t (in general)

pve
28 January 2006 at 10:10

>The need is still there. The analysis shows that a new technology would be needed. Your hypothesis that these developing countries will be the early mass adopters for IPv6 is very valid. By the same reasoning one would expect mobile operators to take the lead.

Another correspondent pointed at http://

Anonymous
30 October 2008 at 17:43

>This is all just really bad pseudo statistics.

Leave a Reply