Download speed test

[NoRulez]

Had a complaint from a colleague in Bloemfontein about slooooww download speeds from a US site where we get self study exam materials so I decided to do a speed test on the different ISP's we use...
Our company uses @lantic as ISP in two branches and MWeb in another branch. I have my own ISP, Telscum/Melkhom/Geldbom, call it whatever you like and the company only pays the telephone call part thereof.
The test criteria was:

• The same 492Mb zip file from the US site with download started within the same hour.
• Only one file downloading at the time.
• No other users on the local ADSL network.
• Monitoring with DU Meter Stopwatch function for first 30 minutes of download.

Here are the results from 4 different offices on 512 ADSL:

@lantic 1: Maximum transfer rate: 212.9 kbps
Average transfer rate: 105.0 kbps

@lantic 2: Maximum transfer rate: 218.7 kbps
Average transfer rate: 119.5 kbps

MWeb : Maximum transfer rate: 389.9 kbps
Average transfer rate: 142.1 kbps

Telkom : Maximum transfer rate: 478.9 kbps
Average transfer rate: 217.0 kbps

Telscum may suck and a lot of other things, but certainly looks like it kicks serious butt!

 

[mancombseepgood]

Had a complaint from a colleague in Bloemfontein about slooooww download speeds from a US site where we get self study exam materials so I decided to do a speed test on the different ISP's we use...
Our company uses @lantic as ISP in two branches and MWeb in another branch. I have my own ISP, Telscum/Melkhom/Geldbom, call it whatever you like and the company only pays the telephone call part thereof.
The test criteria was:

• The same 492Mb zip file from the US site with download started within the same hour.
• Only one file downloading at the time.
• No other users on the local ADSL network.
• Monitoring with DU Meter Stopwatch function for first 30 minutes of download.

Here are the results from 4 different offices on 512 ADSL:

@lantic 1: Maximum transfer rate: 212.9 kbps
Average transfer rate: 105.0 kbps

@lantic 2: Maximum transfer rate: 218.7 kbps
Average transfer rate: 119.5 kbps

MWeb : Maximum transfer rate: 389.9 kbps
Average transfer rate: 142.1 kbps

Telkom : Maximum transfer rate: 478.9 kbps
Average transfer rate: 217.0 kbps

Telscum may suck and a lot of other things, but certainly looks like it kicks serious butt!

Try Verizon then crow!

 

[dabouncer]

Did u use a download accelerator like getright or flashget or similar??

 

[Roman4604]

Since all the connections mentioned use exactly the same network (tho different authentication), the differences are probably a product of location. The performance of the local loop & contention of the local exchange/uplinks is more likely the determining factors.

Why not try an experiment by playing account merry-go-round (swap all accounts between branches) to see if it makes a difference.

 

[NoRulez]

Did u use a download accelerator like getright or flashget or similar??

Accelerators not allowed on the site - single connection per login only.

Since all the connections mentioned use exactly the same network (tho different authentication), the differences are probably a product of location. The performance of the local loop & contention of the local exchange/uplinks is more likely the determining factors.

Why not try an experiment by playing account merry-go-round (swap all accounts between branches) to see if it makes a difference.

4 Towns involved - @lantic is Bloemfontein & Rustenburg, MWeb in Pretoria and Telkom is the Kimberley connection.
Will swap around and do test again...

 

[paxron]

I work in the Internet field and I have met lots of people who use Telkom as their ISP and have better download speeds than myself, their downloads speeds on a 384K connection often reaches above the 40 Kbytes/sec mark, where others would get 38K and below.

 

[mancombseepgood]

Accelerators not allowed on the site - single connection per login only.


4 Towns involved - @lantic is Bloemfontein & Rustenburg, MWeb in Pretoria and Telkom is the Kimberley connection.
Will swap around and do test again...

Are you testing from those locations? If not, then your tests are not giving you what you expect, as someone pointed out, it's only authentication that differs.