Suddenlink → [Speed] DSLReport Tests vs Speedtest.net....

Why is there such a great difference when one tests his speed on this site (specifically, to NJ) and speedtest.net ?
I realize NJ is a little far, but I also notice that's the site most (Suddenlink) customers use for their tests here.
I actually received a sub-one gigabyte download a few minutes ago, but my upload speed was amazingly almost 2 GB. At speedtest.net, my download speed was hitting right at 30mb down, over 2MB up, about what it's supposed to be.
Yes, I do have good, clean connections, and my wireless modem has never affected my speeds so adversely that I've run them straight through the cable modem. I'm actually not really complaining, as iso downloads are plenty fast, as are other downloads, but I did think it strange when I watched a movie earlier today, and the speed was so slow that the stream was buffered for a few seconds.
Suddenlink has provided excellent service overall, and I've had no real problem with the speed in actual apps where speed actually mattered (that movie was the first time I've seen anything like that); I just think this is a strange problem and even though almost all of those sites listed here at dslreports are quite a distance away, speeds should still be in a decent range, not lower than 1/2 of the 30MB down I'm receiving.

I don't think you were downloading/uploading anything anywhere near the gigabyte speed you mention.

That said, dslr speed tests have ALWAYS been slow for a vast majority of people. No idea why. I would recommend just avoiding the speed tests here. Other tools seem to function as you'd expect though.

this is more accurate:

»www.linode.com/speedtest/

Its how close it is, and how its routed (physically and protocol)
I go to speakeasy.net/speedtest a ookla (same as speedtest.net) except they have regular servers vs "choosing the closest one" of course closer will always be better there is traffic shaping occurring everywhere esp when it leaves one network or teir level to another let alone server ability and load factors. or even your own pc with scripts as testmy.net touts that script based tests are notoriously false.
of course there are protocol factors etc when it comes to it.

In my instance speakeasy.net/speedtest have 8 cities to choose from so only 8 servers
Chicago Is regularly at the speed I pay for, and the others are regularly slower than that
even though dallas is physically closer it is about 80% speed others are consistently what they are at but Washington is 33%

I just like using it because I connect to same servers and I get consistent real world results. I dont' expect full speed and low ping times from halfway around the world.

I personally dispise all speedtest sites. I find them to be rarely accurate and are only good for a quick and dirty "ballpark" reading if you are lucky and test when they're running properly. They are either slow or show faster than what your connection will actually sustain due to bursting.

The best "real" test is to find a high throughput server that is as close and low latent as possible. Then stack up downloads and watch a utilization meter like windows task manager (network tab) and see what it does.