You are on page 1of 32

Volume 3, Number 1

The State of
the Internet
1st Quarter, 2010 Report
The “spinning globe” featured in the Akamai NOCC represents where Akamai servers are located and how much traffic they are seeing.
Executive Summary
Each quarter, Akamai publishes a “State of the Internet” report. This report
includes data gathered from across Akamai’s global server network about attack
traffic, broadband adoption, and mobile connectivity, as well as trends seen in
this data over time. Periodically, it also aggregates publicly available news and
information about notable events seen throughout the quarter, including Denial
of Service attacks, Web site hacks, and network events, including outages and
new connections.
During the first quarter of 2010, Akamai observed attack traffic originating from 198 unique countries around
the world. Russia remained the top attack traffic source, accounting for 12% of observed attack traffic in total.
The United States and China once again held the second and third place spots respectively, accounting for nearly
20% of observed attack traffic. Attack traffic concentration returned to levels seen in the third quarter of 2009,
with the top 10 ports once again seeing nearly 95% of the observed attack traffic. We noted that when aggregat-
ed at a continental level, Europe was responsible for the highest percentage of attacks seen in the first quarter,
both overall, and for attacks observed to have originated in known mobile networks. Port 445 continued to be
the most highly targeted port for observed attacks, again both overall and for attack traffic originating in known
mobile networks.

Akamai observed a 7.2% increase (from the fourth quarter of 2009) globally in the number of unique IP addresses
connecting to Akamai’s network. From a global connection speed perspective, South Korea continued to have
the highest level of “high broadband” (>5 Mbps) connectivity. South Korea also maintained the highest average
connection speed, at 12 Mbps, and recorded the highest average maximum connection speed, at 33 Mbps, where
the per-IP address maximum connection speed was averaged across IP addresses from each country. Cities in South
Korea also held many of the top spots in the rankings of highest average and average maximum1 connection
speeds by city. In the United States, Delaware remained in the top position, with 71% of connections to Akamai
occurring at 5 Mbps or greater. Delaware also maintained the highest average connection speed in the United
States, increasing to 7.6 Mbps, and recorded the highest average maximum1 connection speed across the
United States, at 25 Mbps.

In the first quarter of 2010, average measured connection speeds on mobile network providers around the
world ranged from 7.2 Mbps, down to 105 Kbps – both were observed on mobile providers in Slovakia. Of the
109 mobile network providers listed in the report, 35 achieved average connection speeds above 1 Mbps. Average
maximum1 connection speeds on mobile providers around the world ranged from over 20 Mbps down to just over
400 Kbps. Of the 109 listed providers, 83 achieved average maximum1 speeds greater than the 2 Mbps broadband
threshold, 33 achieved average maximum1 speeds greater than the 5 Mbps high broadband threshold, and six
achieved average maximum1 speeds greater than 10 Mbps.
Table of Contents
Introduction 5
Section 2: Security 6
2.1 Attack Traffic, Top Originating Countries 6
2.2 Attack Traffic, Top Ports 7
Section 3: Internet Penetration 8
Section 4: Geography– Global 9
4.1 Global Average Connection Speeds 10
4.2 Global Average Connection Speeds, City View 11
4.3 Global Average Maximum Connection Speeds 13
4.4 Global Average Maximum Connection Speeds, City View 14
4.5 Global High Broadband Connectivity 16
4.6 Global High Broadband Connectivity: Speed Distribution 17
4.7 Global Broadband Connectivity 18
4.8 Global Narrowband Connectivity 19
Section 5: Geography– United States 20
5.1 United States Average Connection Speeds 20
5.2 United States Average Connection Speeds, City View 21
5.3 United States Average Maximum Connection Speeds 21
5.4 United States Average Maximum Connection Speeds, City View 22
5.5 United States High Broadband Connectivity 22
5.6 United States High Broadband Connectivity: Speed Distribution 23
5.7 United States Broadband Connectivity 24
5.8 United States Narrowband Connectivity 24
Section 6: Mobile 25
Section 7: Appendix 28
Section 8: Endnotes 29

4 © 2010 Akamai Technologies, Inc. All Rights Reserved


Introduction
Akamai’s globally distributed network of servers allows us to gather massive
amounts of information on many metrics, including connection speeds, attack
traffic, and network connectivity/availability/latency problems, as well as traffic
patterns on leading Web sites.
In the first quarter of 2010, observed attack traffic continued to target a relatively consistent set of ports,
and attacks targeting port 445 continued to be responsible for an overwhelming percentage of the observed
attacks, both overall, and for attacks observed to be originating in known mobile networks. Russia remained
the top source of observed attack traffic, followed closely by the United States – in aggregate, the two countries
were responsible for 22% of observed attack traffic, with the balance originating in a long tail of 196 more
countries/regions. Aggregated at a continental level, Europe is responsible for the highest percentages
of observed attacks, both overall, and for attacks observed to be originating in known mobile networks.

Akamai noted accelerated quarterly growth in the global observed unique IP count in the first quarter of
2010, which was up over 7% from the prior quarter, to over 487 million unique IP addresses making requests
to Akamai. Quarterly trending in global average connection speeds was mixed during the first quarter, though
quarterly trending in average maximum1 connection speeds was generally positive. Given that data from known
mobile networks was removed from the data set used to calculate connection speed metrics, more countries
than expected showed declining quarterly levels of high broadband and broadband adoption during the first
quarter, though we believe that this may be related to the increased consumption of rate-limited streams
delivered over HTTP.2 From a year-over-year perspective, trending was more generally positive for both metrics.
Rates of narrowband adoption also continued to show surprising levels of growth on a global basis.

In the United States, Delaware continued to maintain its position as the state with the strongest connection
speed metrics, placing first for average connection speed (7.6 Mbps) and average maximum1 connection speed
(25 Mbps), as well as high broadband (71%) and broadband (97%) adoption.

In response to the growing amount of Internet content being accessed through mobile devices such as
smartphones and laptops equipped with mobile broadband connection technologies, and also in response
to multiple inquiries for such data, Akamai has begun publishing insights into metrics collected from connections
to Akamai that have been identified as coming from networks associated with mobile providers. During the first
quarter of 2010, average measured connection speeds on mobile network providers around the world ranged
from 7.2 Mbps, down to 105 Kbps – both were observed on mobile providers in Slovakia. Average maximum1
connection speeds on mobile providers around the world ranged from over 20 Mbps down to just over 400
Kbps. Consumption of content from Akamai by users on known mobile networks ranged from 9.5 GB per
unique IP address per month down to just 11 MB per unique IP address per month.

© 2010 Akamai Technologies, Inc. All Rights Reserved 5


Section 2:
Security

Akamai maintains a distributed set of agents deployed across the Internet that
serve to monitor attack traffic. Based on the data collected by these agents, Akamai
is able to identify the top countries from which attack traffic originates, as well as
the top ports targeted by these attacks. (Ports are network layer protocol identifiers.)
This section, provides insight into Internet attack traffic, as observed and measured
by Akamai, during the first quarter of 2010. While some quarter-over-quarter
trending may be discussed, it is expected that both the top countries and top
ports will experience some change on a quarterly basis.
2.1 Attack Traffic, Top Originating Countries Aggregating the observed attack traffic at a continental
During the first quarter of 2010, Akamai observed attack level, we find that Europe is far and away responsible for
traffic originating from 198 countries, the same as in the the largest percentage of attacks seen in the first quarter,
fourth quarter of 2009. For the third consecutive quarter, as shown in Figure 2. In addition, in aggregating observed
Russia held the top spot, originating 12% of observed attack traffic that originated from known mobile networks,
attack traffic, as shown in Figure 1. The United States as shown in Figure 3, we find that the largest percentage
remained in second place though the portion of global of it (approximately half) appears to have come from
observed attack traffic the country generated declined European mobile networks, followed by nearly a third
from 12% to 10% quarter over quarter. India and of it from mobile networks across South America. Given
Argentina dropped out of the top 10 in the first quarter the ports targeted by attacks identified as coming from
(to 13th and 11th places, respectively), while Japan and mobile networks, and the similarity to the overall target
Poland moved into the top 10 (after ranking 11th and port list, we believe that this attack traffic is likely being
12th respectively in the fourth quarter of 2009). Attack generated by infected PC-type clients connecting to wire-
concentration among the top 10 countries declined less networks through mobile broadband connection
slightly from the prior quarter, accounting for 61% technologies, and not by infected smartphones or
of observed attacks. similar mobile devices.

Country/Region % Traffic Q4 ‘09 % 7 10 8


1 Russia 12% 13%
2 United States 10% 12%
3 China 9.1% 7.5%
6 1
4 Taiwan 6.1% 5.5%
5 Brazil 6.0% 6.4% 9
6 Italy 4.4% 4.5% 3
2
7 Germany 3.9% 4.4% 4
8 Romania 3.2% 3.0% 5
9 Japan 2.9% 2.9%
10 Poland 2.4% 2.2%
– Other 39% 37%

Figure 1: Attack Traffic, Top Originating Countries/Regions

6 © 2010 Akamai Technologies, Inc. All Rights Reserved


Africa 1% Microsoft SQL Server 0.9% VNC Server 0.9%
Remote Administrator 1.5% SOCKS Proxy 0.5%
WWW 1.7% Other
South Microsoft-RPC 2.5% 5.6%
America Telnet 2.5%
11%
NetBIOS 3.2%
North
America 13%
SSH 6.3%
Europe 44%
Microsoft-DS
74%

Asia Pacific
31%

Figure 2: Q1 2010 Observed Attack Traffic, Aggregated by Region Port Port Use % Traffic Q4 ‘09 %
445 Microsoft-DS 74% 74%
22 SSH 6.3% 5.2%
139 NetBIOS 3.2% 2.8%
North America 2.2% Africa 0.1%
23 Telnet 2.5% 2.5%
135 Microsoft-RPC 2.5% 2.8%
80 WWW 1.7% 1.5%
4899 Remote Administrator 1.5% 1.1%
Asia Pacific 1433 Microsoft SQL Server 1.1% 0.9%
16%
5900 VNC Server 0.9% 0.8%
1080 SOCKS Proxy 0.5% 0.3%
Various Other 5.6% –
Europe 50%
Figure 4: Attack Traffic, Top Ports
South
America 32%

the concentration of attacks targeted at this port varies


widely, ranging from 95% in Romania to just over 50%
in the United States. Elsewhere among the top 10, SMTP
(port 25) was replaced in the list by SOCKS Proxy (port
Figure 3: Q1 2010 Observed Attack Traffic from Mobile Networks,
1080), which moved up from 14th place in the fourth
Aggregated by Region
quarter of 2009.

2.2 Attack Traffic, Top Ports In reviewing observed attack traffic originating in known
Attack concentration among the top 10 targeted ports mobile networks, Port 445 was, for most network providers,
increased quarter over quarter, with the top 10 ports far and away the most targeted port, likely indicating user
responsible for just under 95% of observed attacks populations with insufficiently patched systems running
(up from just under 92% in the fourth quarter of 2009). Microsoft Windows OS variants that are connecting to
As usual, port 445 was the most targeted port, once mobile networks, as opposed to malware running on smart-
again targeted by 74% of observed attacks. In looking phones and similar mobile devices. Other popular targets
at the target port distribution across the top 10 countries, for the first quarter included Port 135 (Microsoft-RPC),
Port 139 (NetBIOS), and Port 22 (SSH).

© 2010 Akamai Technologies, Inc. All Rights Reserved 7


Section 3:
Internet Penetration

Through a globally-deployed server network, and by virtue of the billions of requests


for Web content that it services on a daily basis, Akamai has unique visibility into
the levels of Internet penetration around the world. In the first quarter of 2010,
over 487 million unique IP addresses, from 233 countries/regions, connected to the
Akamai network – 7.2% more IP addresses than in the fourth quarter of 2009, and
16% more than in the same quarter a year ago. While the yearly change was roughly
consistent with the level seen in the fourth quarter of 2009, the quarterly change
was nearly 75% higher than that seen in the fourth quarter, possibly indicating
accelerated growth in Internet penetration levels. Although we see approximately
487 million unique IP addresses, Akamai believes that it sees approximately one billion
Web users. This is because in some cases, multiple individuals may be represented by
a single IP address (or small number of IP addresses), as they access the World Wide
Web through a firewall or proxy server. Conversely, individual users can have multiple
IP addresses (handheld, personal/home system, business laptop, etc.).
As shown in Figure 5, the top 10 countries remained Globally, 54 countries saw a quarterly decline in IP address
the same quarter-over-quarter, though Brazil and Spain counts in the first quarter, though nearly two-thirds of those
again exchanged places at the bottom of the list. Con- were places where Akamai observed just tens or hundreds
tinuing the ongoing trend, the United States and China of IP addresses. Concentration among the top 10 remained
accounted for nearly 40% of the observed IP addresses. consistent as well, accounting for approximately 71% of
In addition to the accelerated quarterly growth seen the observed IP addresses for the third consecutive quarter.
in the global IP address count, in comparing fourth In looking at the “long tail,” there were 184 countries/
quarter 2009 and first quarter 2010 quarterly change regions with fewer than one million unique IP addresses
levels, eight of the top 10 countries also saw similarly connecting to Akamai in the first quarter of 2010, 140
accelerated quarterly growth in Internet penetration with fewer than 100,000 unique IP addresses, and 32
levels. (Germany and the United Kingdom were the with fewer than 1,000 unique IP addresses. The count
two exceptions, seeing smaller quarterly changes for 1,000 unique IP addresses remained consistent quarter-
than in the prior quarter.) over-quarter, while the other two were down slightly.

Country/Region Q1 ‘10 Unique QoQ YoY


IP Addresses Change Change 6 4
– Global 487,618,413 7.2% 16%
1 United States 129,354,234 8.1% 11%
5
2 China 57,723,188 14% 30%
3 Japan 33,220,465 3.8% 13%
4 Germany 31,012,378 1.6% 8.6% 8 10 7
5 France 22,473,570 5.1% 15% 2
1 3
6 United Kingdom 20,114,050 1.5% 6.9%
7 South Korea 16,715,485 7.2% 16% 9
8 Canada 11,729,224 3.9% 4.1%
9 Brazil 11,381,433 6.7% 23%
10 Spain 11,224,801 4.8% 13%

Figure 5: Unique IP Addresses Seen By Akamai

8 © 2010 Akamai Technologies, Inc. All Rights Reserved


Section 4:
Geography– Global

Akamai has a unique level of visibility into the connection speeds of end-user
systems, and as such, of broadband adoption around the globe, by virtue of the
billions of requests for Web content that it services on a daily basis through its
globally-deployed server network. Because Akamai has implemented a distributed
network model, deploying servers within edge networks, it can deliver content more
reliably and more consistently at those speeds, in contrast to centralized competitors
that rely on fewer deployments in large data centers. For more information on why
this is possible, please see Akamai’s How Will The Internet Scale? White Paper3
or the video explanation at www.akamai.com/whytheedge.
The data presented within this section was collected for standard-definition DVD quality video content, while
during the first quarter of 2010 through Akamai’s globally- Blu-Ray (1080p) video content has a maximum video bit
deployed server network and includes all countries/regions rate of 40 Mbps, according to the Blu-Ray FAQ.4 As we have
that had more than 1,000 average monthly unique IP done in prior quarters, in order to provide additional insight
addresses make requests to Akamai’s network during the into where users have connection speeds that would allow
first quarter. For purposes of classification in this report, them to be able to effectively consume this higher quality
the “broadband” data included below is for connections media, we will continue to examine how the “high broad-
greater than 2 Mbps, and “high broadband” is for con- band” connections are distributed across speed groupings
nections of 5 Mbps or greater. In contrast, the “narrow- ranging from 5 to >25 Mbps. In addition, starting with this
band” data included below is for connections slower than edition of the State of the Internet report, we will begin
256 Kbps. Note that the percentage changes reflected reporting average maximum1 connection speeds around
below are not additive – they are relative to the prior the world, from a country/region, state, and city perspec-
quarter(s). (That is, a Q4 value of 50% and a Q1 value tive. This metric can provide insight into the peak speeds
of 51% would be reflected here as a 2% change.) A that users can likely expect from their Internet connections.
quarter-over-quarter change is shown within the tables We will also continue to look at which cities around the
in several sections below in an effort to highlight general world have the highest average and average maximum1
trends. A year-over-year change is also shown in some connection speeds.
tables to illustrate longer-term trends.
Finally, starting with this quarter’s State of the Internet
As the quantity of HD-quality media increases over time, report, traffic from known mobile network providers will
and the consumption of that media increases, end users be analyzed and reviewed in a separate section of the
are likely to require ever-increasing amounts of bandwidth. report, and this has been removed from the data set used
A connection speed of 2 Mbps is arguably sufficient for to calculate the metrics reported in the present section.
standard definition TV-quality video content, and 5 Mbps

© 2010 Akamai Technologies, Inc. All Rights Reserved 9


Section 4:
Geography– Global (continued)

4.1 Global Average Connection Speeds With traffic from known mobile networks now removed
After a generally positive showing in Q4 2009, changes from the data set, it is not immediately clear what is causing
to average connection speeds are once again mixed the observed decline in average connection speeds on a
around the world. (Readers may note that the Mbps global basis and in other countries. The base data set may
figures below are higher than those listed in prior editions still include usage from mobile networks not yet identified,
of this report, but are still reflecting quarter-over-quarter or from network providers that are mixing mobile and fixed
declines. As noted in the introduction to this section, network traffic on a single autonomous system. Alternatively,
traffic from known mobile providers has been removed it may point to higher usage of network-reliant applications
from the data sets used to calculate this metric for Q1 and devices that are consuming some amount of network
2010, and the accompanying quarter-over-quarter and bandwidth in communicating with non-Akamai systems.
year-over-year changes have been calculated from this Finally, some of this decline may be related to the delivery
adjusted data set as well.) As shown in Figure 6, average of video content for the 2010 Winter Olympics over the
speed increases among the top 10 countries were largely Akamai HD Network,2 as discussed in Section 4.5 below.
unremarkable, with Hong Kong and Denmark remaining
During the first quarter, 96 countries had average
essentially flat, and Japan, the Netherlands, and Switzer-
connection speeds below 1 Mbps, a level consistent with
land improving by 3.5% or less. Over the long term,
the prior quarter. Akamai measured average connection
year-over-year trends are more positive, however, with
speeds below 100 Kbps in five countries in the first quarter
seven of the top 10 countries having higher average
– up from three in the fourth quarter of 2009. (Note that
connection speeds than during the same period a year
the slowest countries often have the smallest number
ago. The United States ranked 16th globally, managing
of unique IP addresses connecting to Akamai, so it may
slight half-percent increases quarter-over-quarter and
be the case that a few less countries fell below the 1000
year-over year. Given this, the adjusted average connec-
unique IP address threshold in the first quarter than in
tion speed for the fourth quarter would be just under
the fourth quarter.) The lowest average connection speed
4.7 Mbps – nearly a full Mbps higher than the speed
was once again in Mayotte, at 40 Kbps, even with the
reported in last quarter’s report, which included traffic
prior quarter.
from mobile networks.

Country/Region Q1 ‘10 Avg. Mbps QoQ Change YoY Change 9 5


– Global 1.7 -4.7% -1.0%
6
1 South Korea 12 -2.3% 9.6%
7 4
2 Hong Kong 9.0 – 19%
3 Japan 7.9 2.5% -1.6% 1
4 Romania 6.3 -14% 6.5% 10
5 Latvia 6.3 -0.1% 35% 3
6 Sweden 6.2 -0.7% -8.6% 16
8
7 Netherlands 6.0 3.6% 9.6% 2
8 Czech Republic 5.5 -6.4% 0.7%
9 Denmark 5.3 0.1% 6.2%
10 Switzerland 5.3 3.7% -6.1%

16 United States 4.7 0.5% 0.5%

Figure 6: Average Measured Connection Speed by Country/Region

10 © 2010 Akamai Technologies, Inc. All Rights Reserved


4.2 Global Average Connection Speeds, a significant impact on the rankings of cities within the list.
City View The former top three cities – Berkeley (California), Chapel
For the third consecutive quarter, we are once again Hill (North Carolina), and Stanford (California) – have all
examining average measured connection speeds at a city disappeared from the top 100 list, with former 4th place
level. In addition to the 50,000 unique IP addresses “filter” city Masan (South Korea) moving into the top slot. The
that was implemented last quarter for inclusion in the list, city with the highest average connection speed in the
we’ve applied an additional filter for this quarter’s report. United States is now Monterey Park, CA with an average
It was previously noted that cities that were home to one connection speed of 7.2 Mbps.
or more major academic institutions (colleges/universities)
Figure 7 illustrates that Asia dominates the list, with more
featured prominently within the list. To that end, for this
than half (61) of the top cities located in Japan, 12 more in
quarter’s list, connections from known academic networks
South Korea, including eight of the top 10, and Hong Kong.
(autonomous systems) were removed from the source data
In North America, 14 cities made it to the top 100 (12 in the
set in order to mitigate the impact high-speed campus
United States, 2 in Canada) and in Europe, 12 cities across
connections may have had on the overall rankings. As
6 countries reached the top 100.
seen in Figure 8, this new “academic” filter has had

60

12

9
Mbps

0
ay
ea

en

ng
tes

ds

lic
da

y
n

nia

an
pa

rw
or

an

ed

ub

Ko
na
ta

ma

rm
hK
Ja

No
Sw
erl
dS

ep
Ca

ng
Ge
Ro
th

hR
ut

ite

Ho
Ne
So

Un

ec
Cz

Figure 7: Number of Cities in top 100, Average Measured Connection Speed

Cities in Asia dominate the Global Average Connection Speeds top 100 list, with
61 located in Japan, 12 more in South Korea, and Hong Kong. Fourteen cities
from North America made it into the top 100, along with 12 cities in Europe.

© 2010 Akamai Technologies, Inc. All Rights Reserved 11


Section 4:
Geography– Global (continued)

Country/Region City Q1 ‘10 Avg. Kbps Country/Region City Q1 ‘10 Avg. Kbps
1 South Korea Masan 15839 51 Japan Utsunomiya 7885
2 South Korea Poryong 14134 52 Japan Fukui 7801
3 South Korea Kimchon 13845 53 Canada Oakville 7790
4 South Korea Taegu 13796 54 Japan Tokushima 7780
5 South Korea Milyang 13463 55 Japan Yosida 7769
6 South Korea Seocho 12829 56 Japan Hakodate 7761
7 South Korea Ilsan 12500 57 Canada Victoria 7759
8 Japan Usen 12499 58 Germany Baden-Baden 7733
9 Japan Tokai 12377 59 Japan Tokyo 7675
10 South Korea Seoul 11861 60 Japan Yamagata 7659
11 Japan Kanagawa 11541 61 Romania Constanta 7623
12 Japan Shimotsuma 11206 62 Japan Yamaguchi 7521
13 Japan Urawa 11017 63 Netherlands Groningen 7468
14 South Korea Suwon 10736 64 Japan Kochi 7388
15 Japan Asahi 10670 65 Japan Iwaki 7316
16 Japan Tochigi 10248 66 United States Monterey Park, CA 7272
17 Japan Marunouchi 9804 67 Japan Kofu 7266
18 Sweden Umea 9799 68 Japan Okayama 7233
19 Japan Ibaraki 9684 69 Japan Osaka 7135
20 Japan Shizuoka 9402 70 Japan Toyama 7077
21 Japan Nagoya 9262 71 Japan Kagoshima 7044
22 Japan Kyoto 9249 72 South Korea Suyudong 7031
23 Japan Hyogo 9038 73 Japan Saga 6962
24 Japan Yokohama 9001 74 Japan Kumamoto 6943
25 Japan Kobe 8991 75 Japan Tottori 6858
26 Japan Gifu 8975 76 United States Riverside, CA 6845
27 South Korea Yongsan 8950 77 Japan Okidate 6784
28 Japan Sendai 8897 78 Japan Morioka 6776
29 Japan Nagano 8887 79 United States Fairfield, CA 6719
30 Norway Lyse 8887 80 Czech Republic Ceska 6685
31 Japan Chiba 8801 81 United States Capitol Heights, MD 6614
32 Japan Nara 8637 82 Japan Miyazaki 6610
33 Japan Hodogaya 8584 83 Japan Nagasaki 6602
34 Hong Kong Hong Kong 8571 84 Romania Timisoara 6599
35 Japan Wakayama 8559 85 Romania Iasi 6589
36 Japan Fukuoka 8476 86 Japan Akita 6550
37 Japan Yokkaichi 8412 87 United States Hayward, CA 6545
38 Japan Kagawa 8370 88 Sweden Goteborg 6537
39 Japan Niho 8333 89 United States Walnut Creek, CA 6534
40 Japan Soka 8301 90 United States Staten Island, NY 6518
41 Japan Niigata 8248 91 United States Oakland, CA 6476
42 Japan Otsu 8240 92 United States San Mateo, CA 6452
43 Japan Hiroshima 8209 93 Netherlands Joure 6437
44 Netherlands Wageningen 8130 94 United States Olympia, WA 6426
45 South Korea Inchon 8102 95 Japan Oita 6406
46 Japan Hamamatsu 8059 96 Japan Otemachi 6380
47 Japan Matsuyama 8053 97 United States Boston Metro, MA 6378
48 Japan Mito 8044 98 Netherlands Tilburg 6352
49 Japan Kokuryo 8042 99 Japan Sapporo 6317
50 Japan Kanazawa 7999 100 United States Union, NJ 6317

Figure 8: Average Measured Connection Speed, Top Global Cities

12 © 2010 Akamai Technologies, Inc. All Rights Reserved


4.3 Global Average Maximum European countries took six of the remaining slots,
Connection Speeds all with average maximum connection speeds over 15
As mentioned in the introduction to this section, new Mbps, while the United States placed 8th with an average
to the State of the Internet report series this quarter is maximum connection speed of 16 Mbps. Quarterly changes
a review of average maximum connection speeds. This among the top 10 countries were all positive, as were most
metric represents an average of the maximum measured of the yearly changes, with only Sweden and the United
connection speeds across all of the unique IP addresses States showing slight declines. While the 4.7% quarterly
seen by Akamai from a particular geography. The average decline for the global figure is reasonable, representing
is used in order to mitigate the impact of unrepresenta- a slowdown of approximately 300 Kbps, the 24% yearly
tive maximum measured connection speeds. In contrast decline, equivalent to a 2 Mbps slowdown, is surprising.
to the average measured connection speed, the average A review of the source data shows significant yearly average
maximum connection speed metric is more representative maximum connection speed declines in China, India, and
of what many end-user Internet connections are capable a number of South American countries, among others.
of. (This includes the application of so-called speed boost- While the declines in these countries clearly influenced the
ing technologies that may be implemented within the overall global figure, it is not clear what drove the declines.
network by providers, in order to deliver faster download
In looking at the average maximum speed distribution
speeds for some larger files.) Note that data from known
around the world, only South Korea exceeded 30 Mbps,
mobile networks has also been removed from the source
due in no small part to the broad availability of extremely
data set for this metric.
high speed Internet connectivity across the country. Three
As shown in Figure 9, in the first quarter of 2010, more countries/regions registered average maximum
Akamai calculated an average maximum connection speeds in excess of 20 Mbps, while 39 others had average
speed of 33 Mbps in South Korea – almost 3x the average maximum speeds in excess of 10 Mbps. In looking at the
connection speed within the country – with strong growth “high broadband” threshold of 5 Mbps, 58 additional
on a quarterly and yearly basis as well. Asia continues countries exceeded that. Only 12 countries saw average
to lead this metric as well, with South Korea, Hong Kong, maximum connection speeds below 1 Mbps, including
and Japan taking the first three slots in the top 10 list. Mayotte, where the 411 Kbps average maximum speed
was 10x its average speed for the first quarter.

Country/Region Q1 ‘10 Max. Mbps QoQ Change YoY Change 6 4


– Global 6.4 -4.7% -24%
5
1 South Korea 33 15% 28%
2 Hong Kong 30 15% 8.3%
7
3 Japan 26 9.4% 6.8% 1
4 Romania 25 1.8% 13% 9
5 Sweden 19 3.8% -1.5% 3
8
6 Latvia 19 11% 43%
10
7 Belgium 17 8.9% 8.7% 2
8 United States 16 6.7% -1.2%
9 Portugal 16 7.8% 13%
10 Bulgaria 15 32% 63%

Figure 9: Average Maximum Connection Speed by Country/Region

© 2010 Akamai Technologies, Inc. All Rights Reserved 13


Section 4:
Geography– Global (continued)

4.4 Global Average Maximum Connection When viewed at a more granular city level, the average
Speeds, City View maximum1 connection speeds tell an even better story
Following the lead of Sections 4.2 and 4.3 above, starting of the availability of extremely high-speed connectivity
this quarter, the State of the Internet report will also begin around the world. As Figure 11 illustrates, three cities
to review average maximum1 connection speeds at a city had average maximum1 connection speeds in excess
level. As noted in Section 4.2, connections from known of 40 Mbps, while 22 more exceeded 30 Mbps. Nearly
academic networks (autonomous systems) were removed 100 additional cities around the world had average
from the source data set in order to mitigate impact that maximum1 connection speeds above 20 Mbps, and over
high-speed campus connections may have had on the 360 more achieved average maximum1 connection speeds
overall rankings. We’ve also applied a filter of 50,000 between 10 and 20 Mbps. Over 190 more surpassed the
unique IP addresses for inclusion in this list. “high broadband” threshold of 5 Mbps, while over 130
fell between that level and the “broadband” threshold
It comes as no surprise that Japan and South Korea
of 2 Mbps. Only one city (with more than 50,000 unique
rule the top 10 for this metric as well, with six and
IP addresses connecting to Akamai in the first quarter)
four cities respectively among the top 10, as shown
saw an average maximum1 connection speed below
in Figure 11. Additionally, Figure 10 shows that Japan
1 Mbps – Tripoli, Libya.
and South Korea also dominate the top 100 cities, with
Japan holding over half (53) of the spots on the list, and
South Korea taking another 11. In Europe, 8 cities across
five countries are included in the top 100, with four of
those cities in Romania. In North America, the United
States accounts for over a quarter (27) of the entries.

56
52

28
24
Mbps

20
16
12
8
4
0
tes

ay

l
nia
ea
n

ng

ga
y

en
pa

an

rw
or
ta

Ko

rtu
ma

ed
Ja

rm
hK

No
dS

Sw
Po
ng
Ro

Ge
ut
ite

Ho
So
Un

Figure 10: Number of Cities in top 100, Average Maximum Connection Speed

14 © 2010 Akamai Technologies, Inc. All Rights Reserved


Country/Region City Q1 ‘10 Max. Kbps Country/Region City Q1 ‘10 Max. Kbps
1 South Korea Masan 40558 51 Japan Hamamatsu 26265
2 Japan Tokai 40179 52 Japan Fukui 26149
3 Japan Marunouchi 40084 53 Japan Hakodate 25954
4 Japan Kanagawa 38239 54 Japan Iwaki 25844
5 South Korea Kimchon 37976 55 Japan Kofu 25563
6 South Korea Taegu 37420 56 Japan Hiroshima 25280
7 South Korea Poryong 37030 57 United States Monterey Park, CA 25256
8 Japan Usen 36299 58 Japan Yamagata 25218
9 Japan Urawa 36079 59 Japan Tokushima 25156
10 Japan Shimotsuma 35815 60 Japan Tokyo 24752
11 South Korea Milyang 34943 61 Japan Matsuyama 24629
12 Japan Hodogaya 34213 62 United States Federal Way, WA 24628
13 Romania Constanta 34133 63 United States Everett, WA 24588
14 South Korea Seocho 33752 64 United States Olympia, WA 24564
15 Japan Tochigi 33500 65 United States Bellevue, WA 24461
16 Japan Asahi 33372 66 Sweden Umea 24422
17 Japan Soka 33246 67 Japan Yamaguchi 24397
18 Japan Nagano 33045 68 United States Vancouver, WA 24115
19 South Korea Ilsan 32226 69 United States Salem, OR 24013
20 South Korea Seoul 32091 70 Japan Osaka 23942
21 South Korea Suwon 31700 71 Japan Hyogo 23913
22 Japan Chiba 31461 72 United States Hickory, NC 23879
23 Japan Fukuoka 30424 73 Romania Bucharest 23830
24 Japan Yokohama 30137 74 United States Boston Metro, MA 23742
25 Japan Ibaraki 30106 75 Portugal Coimbra 23642
26 Japan Kokuryo 29747 76 Japan Morioka 23572
27 Japan Nagoya 29630 77 Japan Okidate 23533
28 Japan Kobe 29581 78 United States Fairfield, CA 23477
29 South Korea Yongsan 29499 79 United States Beaverton, OR 23316
30 Japan Sendai 29243 80 Japan Toyama 23265
31 Japan Kyoto 29149 81 Japan Sapporo 23261
32 Japan Mito 29129 82 United States Hayward, CA 23107
33 Romania Timisoara 28678 83 Japan Kochi 22911
34 Romania Iasi 28432 84 Japan Okayama 22845
35 Japan Utsunomiya 28332 85 United States Riverside, CA 22803
36 Japan Niigata 28324 86 United States Capitol Heights, MD 22717
37 Hong Kong Hong Kong 28124 87 United States Tacoma, WA 22717
38 Japan Kagawa 27999 88 United States Oakland, CA 22649
39 Japan Gifu 27941 89 United States Oxford, MA 22586
40 Japan Yosida 27898 90 United States Walnut Creek, CA 22509
41 Norway Lyse 27688 91 United States Stone Mountain, GA 22492
42 Japan Yokkaichi 27653 92 United States Aurora, CO 22229
43 Japan Shizuoka 27522 93 United States Union, NJ 22210
44 Japan Niho 27158 94 United States San Mateo, CA 22070
45 Japan Wakayama 26849 95 United States Romeoville, IL 22056
46 Japan Nara 26692 96 Japan Akita 21956
47 South Korea Suyudong 26611 97 United States Staten Island, NY 21898
48 Germany Baden-Baden 26594 98 United States Mishawaka, IN 21796
49 Japan Kanazawa 26519 99 United States Arvada, CO 21742
50 Japan Otsu 26404 100 United States Mount Prospect, IL 21683

Figure 11: Average Maximum Connection Speed, Top Global Cities

© 2010 Akamai Technologies, Inc. All Rights Reserved 15


Section 4:
Geography– Global (continued)

4.5 Global High Broadband Connectivity Overall, we believe that the declines seen in some regions
In the first quarter of 2010, one-fifth of the connections in the first quarter may be related to the streaming of the
to Akamai were at speeds greater than 5 Mbps – down 2010 Winter Olympics on the Akamai HD Network,5 which
slightly from the prior quarter, but a bit more than the leverages HTTP to provide adaptive bitrate streaming, re-
same quarter a year ago. As shown in Figure 12, quarterly sulting in unmatched scale, quality and a highly interactive
changes across the top 10 countries were mixed, though viewer experience. Streams for the Olympics were encoded
none of the changes were significant. However, for most by the broadcasters at six unique bitrates between 350
of the top 10 countries, yearly growth in high broadband Kbps and 3.45 Mbps – all below the “high broadband”
penetration was fairly strong, with seven of them see- threshold of 5 Mbps, with four of the six bitrates below
ing yearly increases greater than 10%. (Note that traffic the “broadband” threshold of 2 Mbps. As such, because
from known mobile networks was removed here as well, delivery of the streams would have been rate-limited to
which may cause some of the metrics to differ significantly those bitrates (due to the encoding rates), this may have
from values published in prior editions of the report.) impacted the percentage of connections to Akamai in
The positive year-over-year changes are an encouraging excess of 5 Mbps. The Akamai HD Network was first an-
trend across the top 10 and on a global basis. In looking nounced in September 2009,6 and its growth in the first
at countries beyond the top 10, those showing year-over- year of availability is a very positive sign, both for customer
year gains outnumbered declines by a 5:4 ratio, though adoption of the service, as well as the continued growth
there were a number of countries that saw significant of video consumption online. Similar to the impact of data
yearly improvements, likely due to relatively low base from mobile networks in prior quarters, and the subse-
levels of high broadband penetration. Similarly, many quent filtering of such data, Akamai plans to implement
of the countries outside the top 10 that saw significant similar filtering of such rate-limited content, as appropriate,
yearly declines also had relatively low base levels of high from future State of the Internet data sets.
broadband penetration.

Country/Region % above 5 Mbps QoQ Change YoY Change 7 6


– Global 20% -5.2% 1.5%
5
1 South Korea 65% -7.6% 25%
8 3
2 Japan 60% 0.2% 4.8%
3 Romania 48% -4.5% 18% 1
4 Hong Kong 45% -9.0% 16% 9
5 Sweden 42% -3.7% -13% 10 2
6 Latvia 41% 2.9% 75% 14
7 Denmark 41% 0.7% 16% 4
8 Netherlands 40% 2.6% 10%
9 Canada 34% 0.9% 47%
10 Belgium 33% -2.9% -1.1%

14 United States 25% -4.3% -2.6%

Figure 12: High Broadband Connectivity, Fastest Countries/Regions

16 © 2010 Akamai Technologies, Inc. All Rights Reserved


4.6 Global High Broadband Connectivity: in the other buckets have increased. For the United States,
Speed Distribution these speed distributions have remained essentially flat,
In an effort to better understand the distribution of as compared to those published in the 1st Quarter, 2009
connections at speeds above 5 Mbps around the world, State of the Internet report.
Akamai has done a more detailed analysis on these
We expect that, on a global basis, as the adoption
connections in order to publish more detailed data on
and rollout of DOCSIS 3.0 technology by cable Internet
the distribution of connection speeds, aggregated into
providers,7 as well as other FTTH initiatives by telecom
5 Mbps ‘buckets,’ as seen in Figure 13.
providers,8 become more widespread the percentage of
In looking back at the analysis published in the 1st connections in higher speed ‘buckets’ will grow over time.
Quarter, 2009 State of the Internet report, we noted The average maximum1 connection speed data presented
that it would be interesting to see if the percentage above demonstrate that these very high speed connections
of connections over 25 Mbps in South Korea continued are certainly available in some areas, but they need to be
to grow. A year later, we observe that while this percent- made more widely available, and priced at levels that more
age has declined slightly (12% to 11%), the percentages subscribers will find affordable.

Country/Region % above 5 Mbps 5–10 Mbps 10–15 Mbps 15–20 Mbps 20–25 Mbps >25 Mbps
1 South Korea 65% 28% 14% 7.7% 4.7% 11%
2 Japan 60% 34% 16% 5.4% 2.1% 2.2%
3 Romania 48% 33% 8.9% 2.6% 1.1% 1.6%
4 Hong Kong 45% 23% 7.0% 4.2% 3.0% 7.4%
5 Sweden 42% 29% 6.9% 2.8% 1.4% 2.5%
6 Latvia 41% 28% 6.6% 2.5% 1.2% 2.3%
7 Denmark 41% 35% 3.9% 0.9% 0.4% 0.7%
8 Netherlands 40% 31% 5.7% 1.4% 0.6% 1.5%
9 Canada 34% 28% 3.7% 0.8% 0.4% 0.8%
10 Belgium 33% 31% 1.3% 0.2% 0.1% 0.3%

14 United States 25% 20% 2.7% 0.9% 0.4% 0.8%

Figure 13: High Broadband Connectivity, Distribution of Speeds

We expect that, on a global basis, as the adoption and rollout of DOCSIS


3.0 technology by cable Internet providers, as well as other FTTH initiatives
by telecom providers, become more widespread the percentage of connections
in higher speed ‘buckets’ will grow over time.

© 2010 Akamai Technologies, Inc. All Rights Reserved 17


Section 4:
Geography– Global (continued)

4.7 Global Broadband Connectivity pushing Switzerland into second place. While Monaco
Similar to what was noted in Section 4.5 above, we certainly has an impressive rate of broadband adoption,
believe that the rate-limited delivery of streaming content it is derived from a much smaller sample set than the
for the 2010 Olympics over the Akamai HD Network2 may balance of countries among the top 10 – it had fewer
have impacted the quarterly and yearly changes for global than 20,000 unique IP addresses that connected to
broadband connectivity globally and in selected countries Akamai at speeds above 2 Mbps, while other countries
around the world, as shown in Figure 14. In addition, as in the top 10 had hundreds of thousands or millions
was noted previously, data from known mobile networks of unique IP addresses connecting to Akamai at those
was removed from the source data set for this metric. speeds, so its ranking must be considered in that light.

Having said that, quarterly changes among the top Overall, broadband adoption around the world continues
10 countries were mixed, though the yearly changes, to be strong and increasing. In the first quarter of 2010,
on the whole, were certainly more positive. Of interest 50 countries/regions had broadband adoption levels in
is Monaco taking the top slot for global broadband, excess of 50% – this is up from 45 countries/regions in the
first quarter of 2009, and just 27 in the first quarter of 2008.

Country/Region % above 2 Mbps QoQ Change YoY Change 7 10 6


– Global 53% -3.6% -4.3%
1 Monaco 92% 1.5% 7.5%
5
2 Switzerland 91% -0.3% –
9
3 Hong Kong 90% -1.4% 2.9% 4
4 South Korea 89% -3.2% 7.9% 2
5 Bulgaria 89% 5.1% 19% 8
6 Latvia 88% 2.1% 40% 41
1
7 Denmark 87% -0.9% 2.9% 3
8 Japan 87% -1.6% -2.7%
9 Belgium 87% -2.6% -3.5%
10 Slovakia 86% -1.4% 4.2%

41 United States 56% -3.3% -9.6%

Figure 14: Broadband Connectivity, Fast Countries/Regions

Analysts at research firm Point Topic have estimated that emerging countries,
including the BRICs, as well as others in Eastern Europe, Southeast Asia, and
South America, will be the main driver of broadband growth over the next five
years, with a 14% annual growth rate in the number of connections. By 2014
they will account for over 320 million connections, 43% of the projected world
total of 740 million by that time.
[http://point-topic.com/content/dslanalysis/BBAfore100301.htm]

18 © 2010 Akamai Technologies, Inc. All Rights Reserved


4.8 Global Narrowband Connectivity world’s slowest countries, as the data in Figure 15 illustrates,
In looking at narrowband connectivity, in contrast to the with many of the countries in the top 10 seeing quarterly
high broadband and broadband rankings, quarterly and increases in levels of narrowband connectivity, and a yearly
yearly declines are considered to be a positive trend, as it increase across all. However, none of the top 10 countries
likely indicates that higher speed connectivity is becoming reported more than 6,000 unique IP addresses connecting
more widely available and more widely adopted. However, to Akamai at narrowband rates (and none with more than
while broadband adoption continues to increase in many 7,800 unique IP addresses overall), which ultimately means
countries around the world, many other countries are that small shifts in IP address counts can equate to large
still stuck with low-speed Internet connections, with large quarterly or yearly changes. Among countries registering
percentages of their connections to Akamai occurring at hundreds of thousands or millions of unique IP addresses
speeds below 256 Kbps. While data from known mobile connecting to Akamai at narrowband speeds, both quar-
network providers was removed from the data set used terly and yearly changes were mixed, with large percentage
to calculate the metrics reported in this section, it appears increases seen in China and several countries in South
that this did not have a significant impact among the America and Europe.

Country/Region % below QoQ YoY


256 Kbps Change Change
– Global 5.2% 18% 17%
1 Mayotte 99% -0.4% 21%
5
2 Wallis And Futuna 98% 1.2% 39%
3 Equatorial Guinea 98% 7.6% 49%
4 Cook Islands 96% 7.7% 88% 7 9 8
5 Cuba 95% 0.4% 31%
118 6
6 Vanuatu 94% 7.3% 25% 4 3 1
7 Guyana 93% -2.6% 48%
8 Ethiopia 93% 1.3% 29% 2
9 Congo 92% 9.4% 74% 10
10 Rwanda 92% 14% 24%

118 United States 4.3% -4.8% 5.7%

Figure 15: Narrowband Connectivity, Slowest Countries/Regions

© 2010 Akamai Technologies, Inc. All Rights Reserved 19


Section 5:
Geography– United States

In prior editions of the State of the Internet report, metrics for the United States
were incorporated into the overall global metrics section. Starting with this edition
of the report, the metrics for specific geographic regions will now be presented
within their own section. The metrics for the United States presented here are based
on a subset of the data used for Section 4, and are subject to the same thresholds
and filters discussed within the prior section. (The subset used for this section
includes connections identified as coming from networks located in the United
States, based on classification by Akamai’s EdgeScape9 geolocation tool.)
5.1 United States Average Connection Speeds gaining 37%, 15%, and 11% respectively. Looking at
The overall average connection speed for the United year-over-year changes, five states showed average speed
States as a whole in the first quarter of 2009 was 4.7 increases of 10% or more, while seven states showed aver-
Mbps. This was exceeded by 22 states, including those age speed declines of 10% or greater. Alaska continued
in the top 10, as shown in Figure 16. Across the country to be the state with the lowest average connection speed,
and within the top 10, most quarterly fluctuations were at 2.7 Mbps, though this is up 30% year-over-year, which
comparatively minor, though Montana, Wyoming, and is certainly a positive sign.
Alaska did show impressive quarterly increases,

State Q1 ‘10 QoQ YoY 4


Avg. Mbps Change Change 8
9
1 Delaware 7.6 -2.0% 4.2% 3
2 District Of Columbia 5.9 -2.3% 3.6% 5
3 Massachusetts 5.9 2.2% 2.5% 10
4 New Hampshire 5.8 3.9% -10% 7
1
5 Rhode Island 5.6 3.4% 1.9%
6
6 Maryland 5.6 -1.1% 1.6% 2
7 Utah 5.6 0.6% -14%
8 Vermont 5.4 -8.2% -3.5%
9 New York 5.4 -1.8% -6.3%
10 Connecticut 5.4 0.2% 3.6%

Figure 16: Average Measured Connection Speed by State

California does have strong broadband connectivity in some locations, although


it ranks 11th among all of the U.S states. Among the more than 250 cities
ranked across the United States, over 25 of them were in California.

20 © 2010 Akamai Technologies, Inc. All Rights Reserved


5.2 United States Average Connection over 25 of them were in California. While New York only
Speeds, City View had a single city within the top 10, 16 cities appeared in the
As with the Global Average Connection Speeds, City overall ranking. Five more states (Texas, New Jersey, Florida,
View presented in Section 4.2, connections from known Michigan, and Virginia) had 10 or more cities in the over-
academic networks were removed from the data set to all list, and over half of the cities listed were concentrated
mitigate the impact that high-speed campus connections across just 10 states.
may have had on the overall rankings. In addition, the
5.3 United States Average Maximum
50,000 unique IP address filter was used for this view
Connection Speeds
as well. In reviewing the top 10 cities in the United States
The overall average maximum1 connection speed calculated
with the highest average connection speeds, as shown
by Akamai for the United States as a whole was 16 Mbps
in Figure 17, it appears that California does have strong
for the first quarter of 2010. This was exceeded by half
broadband connectivity in some locations, although
of the states within the country, including all of those in the
it ranks 11th among all of the U.S states. Among the
top 10, as shown in Figure 18. For this metric, Idaho ranked
more than 250 cities ranked across the United States,
lowest, with an average maximum1 connection speed of
9.3 Mbps, up 6.6% quarter-over-quarter. All but three
City Q1 ‘10 Avg. Kbps states (Wyoming, Arkansas, and Idaho) recorded average
1 Monterey Park, CA 7272 maximum1 connection speeds above 10 Mbps for the first
2 Riverside, CA 6845
quarter. Quarterly trending of average maximum1 connec-
3 Fairfield, CA 6719
4 Capitol Heights, MD 6614 tion speeds within the United States was very strong, with
5 Hayward, CA 6545 only Mississippi and South Dakota heading lower during
6 Walnut Creek, CA 6534 the first quarter, and only just barely, at that. Seven states
7 Staten Island, NY 6518
saw quarterly increases of 10% of more as well. However,
8 Oakland, CA 6476
9 San Mateo, CA 6452 the picture for year-over-year trends was not as bright,
10 Olympia, WA 6426 with only 17 states and the District of Columbia recording
yearly increases, ranging from Colorado at 0.4% to South
Figure 17: Average Measured Connection Speed,
Top United States Cities by Speed Dakota at 28%, and 32 states declining year-over year,
from Nebraska’s 0.1% loss to Idaho’s 17% loss. (Ohio
was flat year-over-year.)

State Q1 ‘10 QoQ YoY 3


Max. Mbps Change Change 7
10
1 Delaware 25 3.4% -5.8% 5
2 Rhode Island 21 7.9% 8.7% 2
3 New Hampshire 21 9.4% -3.4%
4 Hawaii 20 3.2% 27%
8 1
5 Massachusetts 20 7.6% 4.1%
9
6 District Of Columbia 20 3.0% 3.7% 6
7 Vermont 19 0.3% -2.8%
8 California 19 15% 4.4%
9 Maryland 18 8.5% -0.8% 4
10 New York 18 5.6% -7.0%

Figure 18: Average Maximum Connection Speed by State

© 2010 Akamai Technologies, Inc. All Rights Reserved 21


Section 5:
Geography– United States (continued)

5.4 United States Average Maximum 5.5 United States High Broadband Connectivity
Connection Speeds, City View In the first quarter of 2010, Delaware broke its multi-quarter
While California had a strong showing among the top streak of posting significant double-digit quarterly gains in
10 cities for average connection speeds within the United the percentage of connections to Akamai at speeds above
States, the state, on average, was largely displaced by 5 Mbps, delivering a 2.1% quarterly decline, ending the
its neighbors to the north when looking at average quarter with 71% high broadband adoption. Four other
maximum1 connection speeds. Among the top 10 cities states among the top 10 also saw decreasing levels of high
in the United States with the highest average maximum1 broadband adoption quarter-over-quarter, while four states
connection speeds, five of the cities are in Washington, and the District of Columbia saw high broadband adoption
and one is in Oregon, as shown in Figure 19. Concentra- levels grow in the first quarter, as highlighted in Figure 20.
tion of the ranked cities across all of the states was
Across the whole country, 22 states and the District
identical to that seen for average connection speeds,
of Columbia saw quarter-over-quarter increases, from
as discussed in Section 5.2. Interestingly, for the lists
Montana’s doubling to 24% to Virginia’s comparatively
of both average and average maximum connection
small 1.2% growth. Quarterly declines in the remaining
speeds, only Delaware, North Dakota, and Vermont
28 states ranged from Alabama’s 1.8% loss to Nevada’s
did not have any cities on either list. (Given their relative
29% drop. Yearly changes were more balanced, with
rankings on a state level, this likely indicates that cities
25 states increasing, and 25 states and the District of
within those states fell below the 50,000 unique IP
Columbia declining. Alaska and South Dakota posted
address threshold, thereby eliminating them from
the highest levels of yearly growth, increasing 186%
consideration for this metric.)
and 116% respectively. As noted previously, we believe
that the declines observed during the first quarter may
City Q1 ‘10 Max. Kbps be due, at least in part, to delivery of streaming video for
1 Monterey Park, CA 25256
the 2010 Winter Olympics over the Akamai HD Network.2
2 Federal Way, WA 24628
3 Everett, WA 24588
4 Olympia, WA 24564
5 Bellevue, WA 24461
6 Vancouver, WA 24115 Leichtman Research Group, Inc.
7
8
Salem, OR
Hickory, NC
24013
23879
found that the nineteen largest cable
9 Boston Metro, MA 23742 and telephone providers in the US –
10 Fairfield, CA 23477
representing about 93% of the market
Figure 19: Average Maximum Connection Speed, Top United – acquired over 1.4 million net addition-
States Cities by Speed
al high-speed Internet subscribers in the
first quarter of 2010, growing to over
73 million subscribers in total.
[http://www.leichtmanresearch.com/press/051210release.html]

22 © 2010 Akamai Technologies, Inc. All Rights Reserved


State % above QoQ YoY 2
5 Mbps Change Change 9 6 7
1 Delaware 71% -2.1% 15% 10 3
2 New Hampshire 54% 4.7% -7.7% 8
3 Massachusetts 47% 3.0% 7.2% 4
4 Connecticut 42% 2.7% 4.2%
1
5 District Of Columbia 40% 2.1% -0.4%
6 New York 40% -10% -17% 5
7 Vermont 39% -14% -17%
8 Rhode Island 38% 13% -7.9%
9 Indiana 35% -5.4% 21%
10 Pennsylvania 35% -9.2% -0.7%

Figure 20: High Broadband Connectivity, Fastest U.S. States

5.6 United States High Broadband between 10-15 Mbps. Similar to the first quarter of 2009,
Connectivity: Speed Distribution the remaining faster ‘buckets’ struggled to achieve even
Although 9 of the top 10 states with the highest levels 3% of connections in the first quarter of 2010.
of high broadband adoption are on the East Coast
The adoption and rollout of DOCSIS 3.0 technology
(Indiana the lone exception), when we look at the states
by cable Internet providers continued apace in the first
with highest percentages of connections to Akamai at
quarter,10 and Google announced11 an initiative to “build
speeds over 25 Mbps, we find that only three East Coast
and test ultra high-speed broadband networks in a small
states are among the top 10 for that sub-metric. With
number of trial locations across the United States” with
at 3.2% of connections to Akamai over 5 Mbps, Utah
1 Gbps FTTH connections, ultimately placing them in
is well ahead of other states, including those listed in Fig-
competition with incumbent telecom providers. As these
ure 21. In addition, Delaware is well ahead overall, with
initiatives become more widespread, we expect that the
high broadband adoption a full 17% higher than second
percentage of connections in higher speed ‘buckets’ will
place New Hampshire, and double ninth place Indiana.
grow over time. The average maximum1 connection speed
In comparing the speed distributions to the same period data presented above demonstrates that these very high
a year earlier, we find very similar patterns, with the speed connections are certainly available in some areas,
majority of high broadband connections measuring but they need to be made more widely available, and
between 5-10 Mbps, with the next largest grouping priced at levels that more subscribers will find affordable.

State % above 5 Mbps 5–10 Mbps 10–15 Mbps 15–20 Mbps 20–25 Mbps >25 Mbps
– United States 25% 20% 2.7% 0.9% 0.4% 0.8%
1 Delaware 71% 57% 8.6% 2.9% 1.2% 1.5%
2 New Hampshire 54% 47% 4.6% 1.3% 0.4% 0.5%
3 Massachusetts 47% 36% 5.7% 2.0% 0.9% 1.7%
4 Connecticut 42% 34% 4.2% 1.6% 0.8% 1.0%
5 District Of Columbia 40% 29% 5.9% 2.2% 1.1% 2.0%
6 New York 40% 33% 4.5% 1.3% 0.6% 1.1%
7 Vermont 39% 30% 4.8% 1.8% 0.7% 1.3%
8 Rhode Island 38% 30% 4.1% 1.5% 0.7% 1.5%
9 Indiana 35% 27% 3.5% 1.4% 0.8% 1.8%
10 Pennsylvania 35% 29% 3.7% 1.1% 0.5% 0.7%

Figure 21: High Broadband Connectivity, Distribution of Speeds

© 2010 Akamai Technologies, Inc. All Rights Reserved 23


Section 5:
Geography– United States (continued)

5.7 United States Broadband Connectivity 5.8 United States Narrowband Connectivity
Similar to the trend seen in Section 5.5 for high In contrast to the mixed quarterly trending for high
broadband connectivity, quarterly changes across the broadband and broadband adoption levels, all of the top
top 10 states for broadband connectivity were mixed in 10 states with the highest levels of narrowband (<256 Kbps)
the first quarter as well, although they were, by and large, connections saw declines in the first quarter of 2010, as
more muted. Year-over-year changes among the top 10 shown in Figure 23. Across the whole United States, 38
states, as shown in Figure 22, were downward in eight states and the District of Columbia saw lower narrowband
of the top 10 states, with only Delaware and Rhode adoption levels in the first quarter, as compared to levels
Island showing gains, and extremely small gains at that. at the end of 2009. Removal of traffic from known mobile
In looking at the United States as a whole, however, networks from the source data set appears to have had the
quarterly gains and losses were evenly split, although largest impact on this metric, as its impact was most evident
only seven states demonstrated yearly growth in here in prior editions of the State of the Internet report.
broadband adoption.
Unfortunately, things are not quite as positive when
As noted previously, we believe that the declines looking at the changes on a year-over-year basis. Of the
observed during the first quarter may be due, at least top 10 states, seven experienced yearly increases, while
in part, to delivery of streaming video for the 2010 across the country, 27 states saw narrowband adoption
Winter Olympics over the Akamai HD Network.2 increase. Most surprising were the 466% and 116% gains
seen in Nevada and Washington, respectively, especially
2 given Washington’s strong showing in the city views for
8 9 7 6
both average and average maximum1 connection speeds,
10
3 as covered previously in this report.
4
1 5

8 10
7 2
5 6

3
4
State % above QoQ YoY 9
2 Mbps Change Change
1 Delaware 97% -0.2% 0.3%
1
2 New Hampshire 90% 1.4% -0.3%
3 Rhode Island 85% 0.3% 0.2%
4 Connecticut 83% 0.5% -1.5% State % below QoQ YoY
256 Kbps Change Change
5 Hawaii 82% 1.4% -0.7%
1 Alaska 8.0% -9.2% 6.6%
6 Maine 79% -3.0% -7.8%
2 New Jersey 7.9% -4.2% 0.9%
7 Vermont 77% -7.0% -7.6%
3 District Of Columbia 7.4% -14% -15%
8 Michigan 76% 2.5% -0.3%
4 Georgia 6.7% -19% -1.5%
9 New York 75% -4.4% -7.6%
5 Washington 6.6% -9.1% 116%
10 Massachusetts 75% 0.1% -1.1%
6 Missouri 6.2% -9.5% 0.5%
Figure 22: Broadband Connectivity, Fast U.S. States 7 Illinois 6.1% -0.5% 36%
8 Iowa 5.8% -16% 1.6%
9 Texas 5.3% -12% 23%
10 Ohio 4.4% -15% -15%

Figure 23: Narrowband Connectivity, Slowest U.S. States

24 © 2010 Akamai Technologies, Inc. All Rights Reserved


Section 6:
Mobile

Building on the data presented in the State of the Internet reports for the 3rd and
4th quarters of 2009, Akamai continues to identify additional mobile networks
for inclusion in the report, and we continue to expand the amount of information
provided for each network. New to the State of the Internet report in the first quarter
of 2010, we are including information on the average maximum1 connection speed
by provider, as well as insight into the average megabytes (MB) downloaded from
Akamai per month per unique IP address associated with the network. As was noted
last quarter, the source data set for this section is subject to the following constraints:
• A minimum of 1,000 unique IP addresses connecting In examining the data shown in Figure 24, we see that there
to Akamai from the network in the first quarter of is an extremely wide range in average connection speeds –
2010 was required for inclusion in the list. oddly enough, the highest (7175 Kbps) and the lowest (105
Kbps) were both seen on providers in Slovakia. Of the 109
• In countries where Akamai had data for multiple
mobile providers listed, 14 had average connection speeds
network providers, only the top three are listed,
in the broadband (2 Mbps or above) range, while 35 had
based on unique IP address count.
average measured connection speeds of 1 Mbps or more.
• The names of specific mobile network providers have As more providers launch HSPA+12 and HSDPA13 networks,
been anonymized, and providers will be identified as well as networks based on LTE and WiMAX technology,
by a unique ID. we expect that these average speeds will increase in the
future. However, given the current congestion on some
• Data is included only for networks where Akamai
mobile networks, and the rapid increase in the consumption
believes that the entire Autonomous System (AS)
of rich media content on mobile devices, these expected
is mobile – that is, if a network provider mixes traffic
increases may occur over a longer period of time.
from fixed/wireline (DSL, cable, etc.) connections with
traffic from mobile connections on a single network
identifier, that AS was not included in the source
data set.

• Akamai’s EdgeScape database was used for the


continental assignments.

The GSM Association reports that global Mobile Broadband connections roughly
doubled during 2009 to 200 million. By the end of 2010, they estimate this will
reach 342 million global connections, with 120 million in Europe, 116 million
in the Asia Pacific region, and 58 million in North America.
[http://www.gsmworld.com/newsroom/press-releases/2010/4621.htm]

© 2010 Akamai Technologies, Inc. All Rights Reserved 25


Section 6:
Mobile (continued)

Q1 ‘10 Q1 ‘10 Q1 ‘10 Q1 ‘10 Q1 ‘10 Q1 ‘10


Country/Region ID Avg. Max. Avg. MB/ Country/Region ID Avg. Max. Avg. MB/
Kbps Kbps month Kbps Kbps month
Africa
Africa Belgium BE-2 901 2660 33
Egypt EG-1 394 1706 132 Belgium BE-3 200 408 11
Morocco MA-1 412 2929 447 Croatia HR-1 931 3567 58
Nigeria NG-1 242 2962 199 Czech Republic CZ-1 626 2588 69
South Africa ZA-1 465 835 142 Czech Republic CZ-2 415 2024 117
Asia
Asia Czech Republic CZ-3 1320 3561 140
China CN-1 1930 4774 147 Estonia EE-1 611 2775 174
Hong Kong HK-1 2016 8581 397 France FR-1 275 913 48
Hong Kong HK-2 2004 7955 260 France FR-2 1397 4483 630
Indonesia ID-1 218 4924 9501 France FR-3 481 2775 161
Israel IL-1 988 5137 108 Germany DE-1 248 1036 53
Japan JP-1 946 4180 114 Germany DE-2 2507 7931 1049
South Korea KR-1 1495 3029 32 Greece GR-1 909 5068 237
Kuwait KW-1 909 3065 660 Greece GR-2 455 2581 110
Malaysia MY-1 267 1754 176 Hungary HU-1 1145 5315 127
Malaysia MY-2 872 4407 301 Hungary HU-2 1280 5037 72
Malaysia MY-3 531 2651 293 Ireland IE-1 1894 6983 228
Pakistan PK-1 635 4226 494 Ireland IE-2 1076 7331 566
Saudi Arabia SA-1 636 1778 90 Ireland IE-3 978 6830 426
Singapore SG-2 648 5515 119 Italy IT-1 875 4841 275
Singapore SG-3 1282 5923 292 Italy IT-2 1710 6056 252
Sri Lanka LK-1 765 4950 252 Italy IT-3 2783 9889 332
Taiwan TW-1 1032 4332 131 Lithuania LT-1 1203 5516 255
Taiwan TW-2 622 2539 139 Lithuania LT-2 760 3205 185
Thailand TH-1 661 5677 125 Moldova MD-1 730 2858 52
Europe
Europe Moldova MD-2 1269 4907 107
Austria AT-1 2553 10769 122 Netherlands NL-1 803 1758 19
Austria AT-2 1886 6292 1298 Netherlands NL-2 1704 3536 20
Belgium BE-1 2311 7608 282 Norway NO-1 867 3121 67

Figure 24: Average and Average Maximum Connection Speed, Average Megabytes Downloaded per Month by Mobile Provider

In examining the data on average maximum1 connection may be making use of such an architecture.) Akamai
speeds, we see that the gap is even wider than that seen is investigating methods of mitigating the impact of
for average speeds, ranging from a staggering 34584 these gateways/proxies on the source data sets that
Kbps on a provider in the United Kingdom to 408 Kbps will be used for future editions of the State of the
on a provider in Belgium. The Slovakian mobile provider Internet report. Having said that, average maximum1
with the highest average connection speed fared very well connection speeds on mobile networks around the
on the average maximum connection speed measurement world were fairly strong, with 83 of the 109 listed
as well, clocking in at 20394 Kbps. However, it must be providers achieving average maximum1 connection
noted that a number of mobile network providers make speeds greater than the 2 Mbps broadband threshold,
heavy use of mobile gateways and proxies that will result 33 achieving average maximum1 connection speeds
in higher average and average maximum1 speeds being greater than the 5 Mbps high broadband threshold,
calculated by Akamai, as these speeds reflect gateway/ and six achieving average maximum1 connection
proxy-to-Akamai communications rather than mobile speeds greater than 10 Mbps.
device-to-Akamai communications. (These top providers

26 © 2010 Akamai Technologies, Inc. All Rights Reserved


Q1 ‘10 Q1 ‘10 Q1 ‘10 Q1 ‘10 Q1 ‘10 Q1 ‘10
Country/Region ID Avg. Max. Avg. MB/ Country/Region ID Avg. Max. Avg. MB/
Kbps Kbps month Kbps Kbps month
Norway NO-2 1186 3875 70 Mexico MX-2 584 3878 274
Poland PL-1 3444 10298 119 Mexico MX-3 443 4149 353
Poland PL-2 750 2947 38 Netherlands Antilles AN-1 319 1699 180
Poland PL-3 508 2637 135 Nicaragua NI-1 414 2864 216
Portugal PT-1 323 1331 34 Puerto Rico PR-1 2133 8618 2454
Romania RO-1 375 1899 68 United States US-1 845 1912 34
Russia RU-1 4248 13686 138 United States US-2 829 2103 31
Russia RU-2 586 1933 46 United States US-3 979 2496 183
Russia RU-3 498 1570 56 Oceania
Oceania
Slovakia SK-1 105 418 31 Australia AU-1 658 5470 718
Slovakia SK-2 2225 6112 1472 Australia AU-3 928 3785 103
Slovakia SK-3 7175 20394 567 Guam GU-1 378 1729 154
Slovenia SI-1 1074 5514 109 New Caledonia NC-1 426 1730 254
Spain ES-1 1102 6495 276 New Zealand NZ-2 1073 5208 289
Spain ES-2 379 2691 398 South America
South America
Spain ES-3 701 4222 166 Argentina AR-1 233 1838 97
Ukraine UA-1 175 569 35 Argentina AR-2 354 2150 138
United Kingdom UK-1 1043 6647 456 Bolivia BO-1 128 1390 149
United Kingdom UK-2 2065 8613 496 Brazil BR-1 420 2258 123
United Kingdom UK-3 3701 34584 17386 Brazil BR-2 322 1853 108
North
North America
America Chile CL-1 536 3486 371
Canada CA-1 2445 10972 4339 Chile CL-3 379 3442 272
Canada CA-2 728 1902 553 Colombia CO-1 403 3278 146
El Salvador SV-1 469 2755 187 Paraguay PY-1 144 998 87
El Salvador SV-2 704 4930 373 Paraguay PY-2 280 1847 241
El Salvador SV-3 666 3574 547 Uruguay UY-1 485 3127 195
Guatemala GT-1 371 1924 108 Uruguay UY-2 173 1606 78
Guatemala GT-2 453 3859 434 Venezuela VE-1 385 2374 1248

Finally, starting this quarter, we review the average Of these 20 providers, 14 (70%) of them explicitly
amount (MB) of data downloaded from Akamai per advertised support for Apple’s iPhone family of smart-
month per unique IP address seen from the mobile phones, while the remaining six (30%) offer services
network. As Akamai serves content for a number based on next-generation mobile broadband connection
of smartphone makers, we posited that customers technologies such as WiMAX, HSPA+, or HSDPA. Of
of mobile carriers that supported these smartphones the providers where we recorded over 1000 MB/month
would see higher MB/month levels than other carriers. average usage per unique IP address, we found that
In an effort to validate this supposition, we surveyed five of the seven providers supported the Apple iPhone.
the Web sites of the 20 mobile carriers with the highest
On the other end of the spectrum, we found that for
recorded usage levels. (We excluded provider UK-3,
26 mobile providers (of the 109 listed), there was less than
which had ranked highest on the list, due to their
100 MB of data downloaded from Akamai per unique
suspected usage of a mobile gateway architecture,
IP address per month during the first quarter of 2010.
which inflated their calculated per IP address usage.)

© 2010 Akamai Technologies, Inc. All Rights Reserved 27


Section 7:
Appendix

Country/Region % Attack Unique IP Avg. Connection Max. Connection % Above % Above % Below
Traffic Addresses Speed (Kbps) Speed (Kbps) 5 Mbps 2 Mbps 256 Kbps
Europe
Austria 0.3% 2,230,464 3818 11108 19% 63% 1.5%
Belgium 0.1% 3,159,936 4768 16509 33% 87% 1.2%
Czech Republic 0.2% 1,658,849 5482 14131 31% 77% 1.2%
Denmark 0.2% 2,127,420 5330 14059 41% 87% 0.7%
Finland 0.1% 2,342,939 3936 10808 23% 51% 0.8%
France 1.5% 22,473,570 3212 11118 9.5% 69% 0.8%
Germany 3.9% 31,012,378 3877 12948 17% 80% 1.4%
Greece 0.2% 2,037,975 3070 12123 5.5% 68% 1.9%
Iceland 0.0% 123,362 4407 14459 20% 81% –
Ireland 0.2% 1,411,789 4968 12962 11% 54% 2.4%
Italy 4.4% 10,669,453 2919 10448 3.7% 66% 2.2%
Luxembourg 0.0% 161,387 3134 10732 9.1% 69% 1.5%
Netherlands 0.5% 52,302 5992 14459 40% 79% 1.7%
Norway 0.1% 2,381,176 4951 13738 25% 69% 1.5%
Portugal 0.5% 2,155,350 3983 16014 24% 78% 0.7%
Spain 1.3% 11,224,801 2539 9340 4.2% 54% 1.3%
Sweden 0.3% 4,146,254 6181 19209 42% 75% 1.9%
Switzerland 0.2% 2,644,304 5294 14891 27% 91% 0.7%
United Kingdom 1.2% 20,114,050 3812 12346 15% 78% 1.2%
Asia/Pacific
Australia 0.3% 8,400,812 2613 9720 10% 45% 5.4%
China 9.1% 57,723,188 695 2732 0.2% 3.5% 17%
Hong Kong 0.3% 2,189,347 9010 29570 45% 90% 0.5%
India 2.2% 3,966,687 796 4672 0.6% 4.2% 27%
Japan 2.9% 33,220,465 7863 25790 60% 87% 1.5%
Malaysia 0.9% 1,436,465 1067 6081 0.5% 2.8% 11%
New Zealand 0.3% 1,310,127 2912 10671 7.7% 64% 6.3%
Singapore 0.4% 1,721,822 2722 11173 11% 49% 3.2%
South Korea 1.7% 16,715,485 12021 32708 65% 89% 0.3%
Taiwan 6.1% 6,041,143 4334 14053 20% 64% 1.4%
Middle East
Egypt 0.3% 1,067,623 750 4472 0.3% 4.8% 16%
Israel 0.7% 1,836,778 2988 10922 4.2% 62% 0.2%
Kuwait 0.1% 234,633 1392 6229 1.7% 17% 8.7%
Saudi Arabia 0.6% 1,339,271 2189 7213 1.8% 49% 1.2%
Sudan 0.0% 24,360 383 2185 – – 44%
Syria 0.0% 88,098 3034 5911 17% 56% 21%
United Arab Emirates (UAE) 0.2% 739,970 1225 5248 2.8% 13% 7.4%
Latin & South America
Argentina 2.4% 3,620,142 1426 6196 0.5% 16% 8.1%
Brazil 6.0% 11,381,433 1271 5043 2.0% 14% 18%
Chile 0.4% 1,989,659 2202 8896 2.0% 42% 3.1%
Colombia 1.2% 2,319,106 1530 6248 0.3% 22% 6.6%
Mexico 0.4% 7,767,481 1278 5540 0.4% 9.2% 2.1%
Peru 0.6% 616,377 1012 5852 0.5% 5.7% 3.6%
Venezuela 0.3% 1,901,287 705 3217 – 0.7% 13%
North America
Canada 1.5% 11,729,224 4796 14590 34% 80% 2.1%
United States 10% 129,354,234 4684 16207 25% 56% 4.3%

28 © 2010 Akamai Technologies, Inc. All Rights Reserved


Section 8:
Endnotes

1
T he “average maximum connection speed” metric represents an average of the maximum measured connection
speeds across all of the unique IP addresses seen by Akamai from a particular geography. The average is used in
order to mitigate the impact of unrepresentative maximum measured connection speeds. In contrast to the average
measured connection speed, the average maximum connection speed metric is more representative of what many
end-user Internet connections are capable of. (This includes the application of so-called speed boosting technologies that
may be implemented within the network by providers, in order to deliver faster download speeds for some larger files.)
2
T he Akamai HD Network leverages HTTP to provide adaptive bitrate streaming. Streams for the 2010 Winter
Olympics were encoded by the broadcasters at six unique bitrates between 350 Kbps and 3.45 Mbps – all below
the “high broadband” threshold of 5 Mbps, with four of the six bitrates below the “broadband” threshold of 2 Mbps.
As such, because the streams would have been rate-limited to those bitrates (due to the encoding rates), this may have
impacted the percentage of connections to Akamai in excess of 5 Mbps. Similar to the impact of data from mobile
networks in prior quarter, and the subsequent filtering of such data, Akamai plans to implement similar filtering
of such rate-limited content, as appropriate, from future State of the Internet data sets.
3
http://www.akamai.com/dl/whitepapers/How_will_the_internet_scale.pdf
4
http://www.blu-ray.com/faq/
5
http://www.akamai.com/html/misc/hdnetwork.html
6
http://www.akamai.com/html/about/press/releases/2009/press_092909.html
7
http://www.telegeography.com/cu/search.php?search_term=DOCSIS&Submit=Submit
8
http://www.telegeography.com/cu/search.php?search_term=FTTH&Submit=Submit
9
http://www.akamai.com/html/technology/products/edgescape.html
10
http://bit.ly/9mOWFI
11
http://googleblog.blogspot.com/2010/02/think-big-with-gig-our-experimental.html
12
http://en.wikipedia.org/wiki/HSPA%2B
13
http://en.wikipedia.org/wiki/High-Speed_Downlink_Packet_Access

© 2010 Akamai Technologies, Inc. All Rights Reserved 29


The Internet Revolution Continues…At the Fontainebleau Resort, Miami Beach

Your time will be well spent during this revolutionary event. We look forward to seeing you in Miami!
Join Akamai and more than 500 other Internet revolutionaries
for the 3rd Annual Akamai Global Customer Conference.
During our three-day program, we will explore some of the recent challenges and opportunities
that have taken hold and are defining how business is done online. Trends including the growth
of cloud computing models for enterprise-class applications, the adoption of high definition (HD)
video online, the optimization of mobile content for Internet-connected devices, the realization
of secure e-commerce, and the shift of advertising dollars online to follow the migration of
audiences to new media.

With so many technological advancements for leveraging the Internet, as well as public and
private “cloud” infrastructure, it is imperative for today’s online business leaders to have a forum
to discuss these developments with peers from other leading organizations from around the globe.

Visit www.akamai.com/revolution for complete conference details,


and early registration discounts!
The Akamai Difference
Akamai® provides market-leading, cloud-based services for optimizing Web and mobile
content and applications, online HD video, and secure e-commerce. Combining highly-
distributed, energy-efficient computing with intelligent software, Akamai’s global platform
is transforming the cloud into a more viable place to inform, entertain, advertise, transact
and collaborate. To learn how the world’s leading enterprises are optimizing their business
in the cloud, please visit www.akamai.com and follow @Akamai on Twitter.

Acknowledgements
EDITOR: David Belson
CONTRIBUTOR: Jon Thompson
CONTRIBUTOR: Patrick Gilmore
CONTRIBUTOR: Alloysius Gideon
EXECUTIVE EDITOR: Brad Rinklin
EXECUTIVE EDITOR: Tom Leighton

Please send comments, questions, and corrections to stateoftheinternet@akamai.com

Follow @akamai and @akamai_soti on

Akamai | Powering A Better Internet™


For more information, visit www.akamai.com

Akamai Technologies, Inc.

U.S. Headquarters International Offices


8 Cambridge Center Unterfoehring, Germany Bangalore, India
©2010 Akamai Technologies, Inc. All Rights Reserved. Reproduction
Cambridge, MA 02142 Paris, France Sydney, Australia in whole or in part in any form or medium without express written
Tel 617.444.3000 Milan, Italy Beijing, China permission is prohibited. Akamai and the Akamai wave logo
Fax 617.444.3001 London, England Tokyo, Japan are registered trademarks of Akamai Technologies, Inc. Other
U.S. toll-free 877.4AKAMAI Madrid, Spain Seoul, Korea trademarks used herein may be owned by other companies and
are used for descriptive purposes only. Akamai believes that the
(877.425.2624) Stockholm, Sweden Singapore
information in this publication is accurate as of its publication date;
www.akamai.com such information is subject to change without notice.

You might also like