I noted that I got wildly varying numbers on a laptop and an iPhone (there
is also an iPhone app) and the iPhone app doesn't ask for an address. Both
on the same wifi and connection, and the numbers were off by a lot.
And I meant to include examples, but fingers committed the message
before I could stop 'em. Sorry.
PC/mLab:
Download speed: 4150kbps
Upload speed: 2364kpbs
PC/Ookla:
Download speed: 5044kbps
Upload speed: 1120Kbps
iPhone:
Download speed: 1.75Mbps
Upload speed: 0.23Mbps
I've gotten strange stuff each time I've tried their tests. I
particularly like the factor of 10 difference in upload speeds.
... JG
The FCC is probably doing this because US providers generally don't release actual bandwidth, speeds or latency numbers their consumer
customers get. Advertised numbers often don't mean anything. If
providers want to release better data, it might help the FCC understand
the current environment.
Some US providers have published data for their business customer connections and backbones.
Joe Greco wrote:
I've gotten strange stuff each time I've tried their tests. I
particularly like the factor of 10 difference in upload speeds.
... JG
Yeah...these test are algorithm based and rarely accurate! On our 100Mbps Internet connection (which I know handles 100Mbps) best I could get is 10Mbps down and 14Mbps up.
Wish someone would come up with a much better mouse trap. The only test I've ever found to be fairly accurate is iperf or a simple FTP test.
I could imagine that the FCC sees it as a data source.
There are obviously some variables, buffering or something out there
since download speeds do not seem to be very consistent running the
tools several times. I tested three times each with the two engines.
From SATX, TWC/RR:
Ookla
Download Speed 24408 28494 22662 Kbps
Upload Speed 483 492 493 Kbps
Latency 18 18 18 ms
Jitter 2 2 2 ms
MLAB
Download Speed 16854 17630 15780 Kbps
Upload Speed 487 493 493 Kbps
Latency 18 17 17 ms
Jitter 2 1 1 ms
Regards
Jorge