SPEED TEST FLUCTUATIONS

do note... Single thread (testmy) testing VS Multi thread (speedtest.net) testing.

Multithreaded test to Amsterdam:
hGY4VIu.png


Still not impressed :p

Speedtest.net to Amsterdam:

3150876252.png


*** servers is *** :D
 

Optimised servers? Explain?

The servers on Speedtest are independently hosted. The one I used is owned by NFOrce Entertainment in Amsterdam.

There are worse performing ones as well. What would I gain by using a poorly functioning server?

Point is that Testmy is not a reliable indicator of your bandwidth speed--which is ultimately the point of a speed test (which, in itself, also isn't completely accurate because bandwidth speed does fluctuate).

Real-world bandwidth fluctuates heavily depending the servers used as well--not all servers even upload anywhere near 50Mbps.

Actually, another speed test service I "like" is SpeedOf.me.

Only problem there is that the speed fluctuates depending on your browser.

For example, I get my full line-speed if I use Firefox, but half of it if I use Chrome :p
 
Last edited:
Well there is always vimoe's HD to see what is the throughput of the line. Perhaps vimeo + speedtests == better results instead of relying on just one of them. Still... I prefer a single thread test over the multi thread as that will tell me what is the worst I can expect VS a "lab" result. That's the reason I would use testmy over speedtest.

Optimised servers? Explain?

The servers on Speedtest are independently hosted. The one I used is owned by NFOrce Entertainment in Amsterdam.

There are worse performing ones as well. What would I gain by using a poorly functioning server?

Point is that Testmy is not a reliable indicator of your bandwidth speed--which is ultimately the point of a speed test.

Real-world bandwidth fluctuates heavily depending the servers used as well--not all servers even upload anywhere near 50Mbps.

Actually, another speed test service I "like" is SpeedOf.me.

Only problem there is that the speed fluctuates depending on your browser.

For example, I get my full line-speed if I use Firefox, but half of it if I use Chrome :p
 
Well there is always vimoe's HD to see what is the throughput of the line. Perhaps vimeo + speedtests == better results instead of relying on just one of them. Still... I prefer a single thread test over the multi thread as that will tell me what is the worst I can expect VS a "lab" result. That's the reason I would use testmy over speedtest.

Funny thing about HD video, I've never seen it go particularly fast. These videos don't go much faster than 400-600kB/s.

The initial download peaked at 2.6MB/s and then settled down to a few hundred kB/s :)

Give me something more demanding :D
 
Funny thing about HD video, I've never seen it go particularly fast. These videos don't go much faster than 400-600kB/s.

The initial download peaked at 2.6MB/s and then settled down to a few hundred kB/s :)

Give me something more demanding :D

LOL. Well 600kB/s for me is fast :erm: :D
 
The fact that you were on a 2Mb Telkom line and a 1Mb Mweb DSL account meant that you were probably hitting what is know as a burst speed, which is a higher speed for a limited time only. Now that you are upgraded on Mweb side your speed will remain fairly consistent as you are no longer hitting the burst speed.

That being said their could still be speed fluctuations every now and then due to congestion etc. Although by the speeds initially posted I think you are on a fairly uncontested exchange.

This is why I initially asked if the Mweb side was also 2Mb

I think you've got it dead right. Burst speed is what I had. That's why the tests fluctuated between the 1.70's, 1.40's and 1.30's. It is now 42 hours since Mweb put my a\c on 2 Mbps and ALL my tests are consistantly in the 1.70's. It appears that the signal my router receives now from Mweb is steady.
 
Top
Sign up to the MyBroadband newsletter
X