Bandwith Tests

[petedk]

I have been having a few problems recently with my ADSL connection and therefore perform bandwith tests 3 or 4 times a day.

This evening I decided to test again even though I thought the speed was "reasonable". I was surprised to see a speed of 276 kbps. I am using a 528 kbps connection. I tested from 2 other sources (one is Thai Visa) and the speed was 437 and 408. I performed the tests several times from each source over a period of 20 minutes and got pretty much the same results from the same sources.

Why the big difference?

[george]

The tests are ok in my opinion. If you test within Thailand you get the domestic test results only.

If you want to know the nearest test results from outside Thailand, Singapore in our case, you might want to check our own speedtest:

http://speedtest.thaivisa.com

[sjaak327]

to test your own connection, I would also use a speedtest which is geographically closest to your own location,

I just did three speedtest, one to the USA which gave me 4474 down and 566 kbps up, then one to thaivisa, which got me 2665 down and 393 up, finally one to nuria (which is based in the Netherlands, where I am now) and I got a whopping 11831 kbps down and 647 up.

[PattayaAddict]

Dont forget that if you do a test after 10 pm you also tend to get better readings ,happens with my machine all the time .

[Firefoxx]

Speed tests are reliable, in a sense that they can only test within their own parameters. They are accurate in that they:

1. Test the speed from you to their server.

This means that it's not accurate for any other place. It's a good guesstimate of the results to the general country that the server is located in, but not anywhere else.

2. Test the http (web) speed.

ISPs may or may not throttle (or give different priority to) certain protocols, like P2P, so speeds for those protocols may be totally different. Also, sometimes a HTTP proxy may mess up certain speed tests.

3. Test the speed at only one time period.

Speeds fluctuate from hour to hour, minute to minute, day to day, etc. It may be a regular fluctuation (as in good speeds during the night, every night) or it may not.

Speeds tests are a general indicator. If you usually use servers in the country that you do the speed test, then it's fairly valid, taking in points 2 and 3. Me, I've had wildly different results and very steady results, depending on the ISP and time of day/week/month/year.

[jackr]

to test your own connection, I would also use a speedtest which is geographically closest to your own location,

I just did three speedtest, one to the USA which gave me 4474 down and 566 kbps up, then one to thaivisa, which got me 2665 down and 393 up, finally one to nuria (which is based in the Netherlands, where I am now) and I got a whopping 11831 kbps down and 647 up.

You lucky so and so... for your 12meg d/l that is, not for being stuck in tulipland

[Butchersdog]

I am informed that the one on this site is unreliable. On my 1meg i get a test reading here of only 280kbs. I tried www.speedtest. no doubt the like will be removed, however that gave a reading of 840kbs download, and 83kbs upload, which is about right i think.

[cdnvic]

They are accurate to gauge what kind of performance you will get to servers over a similar distance. Thaivisa's server is located in Singapore, and from Canada's west coast the speadtest gives me 1932 down/ 290 up. However, from Seattle (a short distance away) I get 9940 down/ 980 up. The different speeds are an indication of the performance differences due to distances, not the inaccuracy of the tests.