Since tonight I can't seem to sleep, I thought why not test out my DSL connection speed. At home we have two internet connections, one through AT&T, the other through Comcast. Have you ever tested your broadband speed?
I'm posting results from 3 online bandwidth testing sites. I run each test twice and recorded the second value in the table below.
Test 1: The first is a flash-based tool located at http://www.pcpitstop.com/internet/bw.asp, it's really easy to use and fairly simple graphics, and has 4 server options. I was testing against the server in Washington DC.
Test 2: The second is another flash-based tool located at http://www.speedtest.net/, it's much more elaborate graphics, and has a lot of different server options. This was testing against a server in Southfield, MI.
Test 3: The third test is another flash-based tool, located at http://www.dslreports.com/speedtest?flash=1. I trusted DSLreports.com more than other sites, as I have read articles there before. It has 6 server options. I chose Chicago, IL for my first test. The interface shows an analog meter, but the flash applet doesn't run very smoothly (I can see it freeze up and the results get skewed as soon as it does). The results are pretty far off from what the other tests have reported. That is to a sprint server, so at first I thought that might be the issue, but it gives equally skewed results with Washinton DC or Toronto. Note that they're also selling/referring broadband plans. Undue bias? This one gives the error "
Test 4: The fourth test is http://testmy.net/. They're trying hard to get me to run extra tests on my computer to explain why the results are so slow. The exact text was: "I see that you're on a Windows PC. Test PC Speed after your internet test to discover hidden issues with your computer." Bias perhaps?
Test 5: The fifth test is http://reviews.cnet.com/internet-speed-test/. Somehow, I get much better results with the CNET test, again closer to the first two results. Too bad there are no upload or ping results.
I'm testing two connections: our home AT&T connection (lowest cost possible), and CrossTown Wireless (running through Comcast), which we get for free through our apartment complex. It's definitely worth mentioning that I get a 5-bar signal strength on a 802.11n WPA-PSK network connection for my home AT&T, while the CrossTown Wireless is through a 2-3 bar 802.11g unsecured connection. It's definitely not a fair comparison.
Well, the test results seem pretty clear, particularly in light of me using a far worse wireless signal with the Comcast tests. Now I have to explain to myself why we pay for AT&T service (we like having our own connection). On the positive side, my paid internet is apparently supposed to be even slower (for downloading). Perhaps our situation would be different if our neighbors were sucking up tons of bandwidth watching TV shows or downloading torrents on the apartment complex connection. At the very least, I think I'll start watching TV online through our free Comcast connection.