Why is VM Broadband different?

[spot-wobble]

On the same subject,

Can anyone make sense of the download/upload speeds registered by this test?

The tests were carried out using an Apple iMac connected through WIFI to a standard Orange (formerly France Telecom) router which connects to the network via copper.

Andy

[Terry]

Can anyone make sense of the download/upload speeds registered by this test?

Download speed 69Mbps and upload speed 17.5Mbps it says.

What is it that you don't understand about it?

Here's my VM result:

   

This is a nominal 100Mbps service but it's early evening and demand is increasing! (See earlier comments in this thread.)

Also, VM's upload speeds are limited compared with other methods of delivery - also previously discussed, so can't hope to match the speed you are getting.

Interestingly, if I force the test to use VM's own Nottingham server, the speeds are nigh on identical - pointing to the local network as being the limiting factor, which I can well understand - although the ping time is halved because Nottingham is much closer than Sheffield.

[spot-wobble]

Terry,

Previous tests in Scotland used to show a download speed of around 5Mbps and an upload speed of around 320Kbps.
So what has changed?

Andy

[Terry]

Well ...

You aren't in Scotland any more, for a start ...

The way you get your internet has obviously changed - very much for the better, it would appear!

Presumably your previous service was ADSL on a (long) bit of copper wire.

Presumably your new service is FTTC and a short length of wire!

Didn't you check what you were paying for when you signed up for the current service?

You'd have definitely had a shock if the change had been reversed!

[ppppenguin]

Andy, unless you're on a service that's meant to be faster than 80/20 I'd say you've nothing to worry or complain about. SInce you're using wifi you may even find the down speed is being limited by that.

[Mike Watterson]

Never connect via wifi for speed tests!

[ppppenguin]

Mike is correct except that in a few circumstances a fast internet connection can max out 100Mb/s wired internet while newer wifi technologies can exceed this. The moral is to use 1Gb/s ethernet when testing fast connections.

[Mike Watterson]

250Mbps WiFi can easily run at 1Mbps.
It's shared. The speed drops more than 1/N as N users added
The speed drops more than 1/4 for every double distance.
The speed can be 1/2 to 1/20th if someone near by fires up on same channel, not just WiFi, but baby alarms, video senders and more, on both bands as they are not dedicated to WiFi. Bluetooth and proprietary wireless mouse/keyboard reduces speed.
Wireless HDMI.

Also typical WiFi uses 4 to all the channels in a band to get more speed. Hence trying a different channel can be pointless. I use a scanning utility with Laptop disconnected to see what is in use and signal levels. It only detects actual WiFi. I have an ancient HP141T with an 18GHz plug-in, and external "aerials" for the GSM, 3G, MMDS/LTE, WiFi (both bands), X-Band Ku, so I can see what is really happening. I use a different plug-in for Radio & TV bands.

[ppppenguin]

As Mike correctly states, there are many things that will slow down wifi. I was making the point that under some conditions wifi can be faster than "vanilla" 100Mb/s ethernet and that 100Mb/s is now inadequate or marginal for some fast internet services.

[spot-wobble]

The only reason the speed test was done with a wifi connected computer is that at the moment that is the only connection we have.
I thought it better to install and earth connection for the house first rather than start pulling cat5 cable around the place!.

As for the service at sign up, I think we did pretty well to get signed up at all at the Orange store in Sarlat given our level of spoken French!
I could try connecting the laptop to the router and trying the test from there...

Andy