Cable vs. DSL: Latency

Home	Blog	Bio	Projects	Contact	Latest Blog (new site): How to Get to Genius

JamesThornton.com $-\>$ Writing $-\>$ Cable vs. DSL: Latency

Cable vs. DSL: Latency

Summary: Compares latency between cable and DSL connections.

I currently use DSL for my Internet connectivity, but I have used cable. My roommate and I both have mobile phones, we never use the house phone, but SBC requires us to pay for a voice line with our DSL connection. We are about to move to a new house, and my roommate has been asking me about switching to cable so that we don't have to pay for a phone line nobody is using.

My roommate argues that cable is faster, we don't have to pay for a phone line, and he can get several free months through Dell where he works in corporate sales. I am an Internet consultant, and I often use a terminal to log in to remote servers with SSH to write code or administer systems. My experience with cable has been miserable because the latency is so bad.

Latency is how long it takes a packet to make a round trip on the network (in simplistic terms, a packet is something that carries data across a network). Bandwidth is the size of the network connection. Together, latency and bandwidth determine the speed of a connection.

For me, high latency causes severe productivity problems because when I am connected to a remote server writing code (which is frequently), and I type a character, the character may not appear on the screen for a few seconds. If the latency is inconsistent, the number of seconds between when I type a character and when it displays on the screen fluctuates. This will drive you crazy, especially if you know that you won't have any of these problems with DSL.

Since my roommate doesn't feel my pain, I ran some tests to show him the difference in latency between our Austin SBC Yahoo! DSL connection and some random Austin RoadRunner cable connection. I ran these tests from my Dallas server, and while they are not scientific, the results are representative of my previous experience with cable and DSL.

DSL: adsl-68-92-149-68.dsl.austtx.swbell.net --- 68.92.149.68 ping statistics --- 333 packets transmitted, 333 packets received, 0% packet loss round-trip min/avg/max/mdev = 20.570/22.597/41.045/1.423 ms

Cable: cs666838-214.austin.rr.com --- 66.68.38.214 ping statistics --- 333 packets transmitted, 333 packets received, 0% packet loss round-trip min/avg/max/mdev = 69.617/80.451/120.461/12.282 ms

Both tests sent and received 333 packets, with no packet loss. Look at the last line of each test, labeled "round-trip". It shows the minimum time (min), average time (avg), maximum time (max), and standard deviation (mdev) of how long it takes to send and receive a packet.

Compare the maximum DSL time (41.045) to the minimum RoadRunner time (69.617).

Now compare the average DSL time (22.597) to average RoadRunner time (80.451), and you'll see that RoadRunner is almost four times slower on average.

Now look at the standard deviation (mdev), and you'll see that the DSL connection has a relatively small standard deviation (1.423) compared to the RoadRunner connection (12.282) so this indicates that RoadRunner's latency is inconsistent.

The inconsistent times are expected with the cable network party-line architecture. With cable, everyone in your area uses the same large connection so if many people in your area are online at the same time, it will be more congested. I ran this test at 1:30 p.m. when most people are at work and not using their cable modem so it's a good bet that the latency and inconsistency will be much worse in the evening.

I have used both DSL and cable, and low DSL latency vs. high and inconsistent cable latency is the reason I use DSL.

Related Resources

It's the Latency, Stupid, by Stuart Cheshire
Comparison of Latency and Bandwidth, from Wikipedia