Speedtest Wildy Varying Results

[bluesofa]

I'm in Rayong with a newly installed 6Mb/1Mb (alledged) TOT ADSL

Today the net seemed pretty slow in the morning, this afternoon was better.

Out of interest I tested with some speedtest sites, all using the Ookla system.

Just talking about speeds within Thailand here.

3BB broadband (TT&T) showed 5Mbd/l, 500K u/l

speedtest.dot.net showed 5Mb d/l, 800K u/l

thaisvisa.dot.com showed 800K d/l, 40K u/l

What I can't understand is that speedtest says it was testing using the thaivsa site in BKK.

I ran these tests 4 or 5 times each, so they are a fair indication of consistant speeds, but why does thaivisa always seem to give a much slower result, and yet speedtest use their Ookla site?

I'm just ocinfused with this inconsistancy.

Cheers,

[siamect]

I was just going to ask you if you used the BKK server or the Singapore server. But when i tested it turns out I get about 6Mbps on both Singapore and BKK server and I get the same on speedtest.net

I have True 12Mbps in and 1Mbps out. So I'm on only 50% on what I would have if I was alone.

Martin

[negreanu]

Actually on speedtest.net the "Hosted by" Thaivisa server always gives slower results to the other (2?) hosted Thailand servers.

[negreanu]

Actually on speedtest.net the "Hosted by" Thaivisa server always gives slower results to the other (2?) hosted Thailand servers.

Tonight on 16mb True Premium 19:45pm speedtest.net results:

SF server 15.31mb

BKK Server (STS Group) 14.73mb Second Run 14.45mb

BKK Server (Thaivisa) 11.22mb Second Run 12.2mb

SIN Server 14.96mb

Edited May 29, 2010 by negreanu

[Pib]

The Ookla speed test is the front end/software/java script being used, but the underlying test server can vary depending on the ISP you are running the test against. Since you are on TOT recommend you use TOT's Ookla front end speed meter at http://www.totisp.net/ as this will probably give you the most accurate speed results from your residence to the TOT DSLAM/central office.

And a key factor of any speed test is the capability of the test/other-end-server which is of course also affected by the bandwidth your ISP has to that location/server, routing/pings to reach that location/server, capability/capacity of that location/server, etc...etc..etc. When you do a speed test to say two different locations somewhere on this world which are phyically pretty close to each other as the crow flies but you get significantly different speed/bandwidth results, probably the biggest reason is the capability/capacity of that test server.

Only use speedtest results as one indicator of whether your internet connection is working as it should. Also use your general browsing snappiness, how long downloads take, email pop3 connection/download, etc., also need to be used in your evaluation.

[hml367]

There are a lot of variables when running so many speedtests. Even the browser you are using at the time makes a difference. I have been given an indication of more than 13 mb download speed to Bangkok when using Google Chrome on AIS Super 3G in Chiang Mai. My most consistent results are using IE.

Still, just too many variables. As indivated in someone's previous post, the speed of a download is the best indication. Even then it depends on the routing between the Thai gateway and the end server.

McAfee has a speedtest which actually downloads a file to give its results.

MSPain