Well, usually the problem is the opposite, but I just ran three different speed tests and got speeds from 25 to 370 Mbps It seems like absolute madness but there is even a certain logic to it, in that the fastest test (Speed 2) is against a local server at CMU, only a few kilometres away and the other tests are with servers in Bangkok. I'm on TOT's platinumcyber, i.e. supposedly 3 Mbps

Can anyone explain these measurements:

/ Priceless

must be a test script error, the speed is not real at all.

looks like the file was tested in lan not wan.

That was tested within ISP/server cache. That happens from time to time. I recently had 30mb reported...

Not necessarily a script error. Nothing to worry about.

That was tested within ISP/server cache. That happens from time to time. I recently had 30mb reported...

Not necessarily a script error. Nothing to worry about.

I believe Priceless ran that speedtest from my university server. I was just about to look at the code for an error until you mentioned the ISP cache which I'd forgotten about. My EDGE test was 120kbs down, 57kbs up on the same server. Daytime speeds, night time about 80% faster.

That was tested within ISP/server cache. That happens from time to time. I recently had 30mb reported...

Not necessarily a script error. Nothing to worry about.

I believe Priceless ran that speedtest from my university server. I was just about to look at the code for an error until you mentioned the ISP cache which I'd forgotten about. My EDGE test was 120kbs down, 57kbs up on the same server. Daytime speeds, night time about 80% faster.

That I was waiting for.... now I feel better

That was tested within ISP/server cache. That happens from time to time. I recently had 30mb reported...

Not necessarily a script error. Nothing to worry about.

I believe Priceless ran that speedtest from my university server. I was just about to look at the code for an error until you mentioned the ISP cache which I'd forgotten about. My EDGE test was 120kbs down, 57kbs up on the same server. Daytime speeds, night time about 80% faster.

You're quite right, Tywais, I used your university server (for the second test, that is). I was thinking about an ISP cache myself, but as far as I can understand (and I am not an engineer) that would eliminate the time from the local server to the server in Bangkok.. What I still don't understand is the speed from my computer to the caching server over my lowly 3 Mbps second ADSL line, and for that matter over my wireless LAN.

Oh well, I guess that I'll file it under "Random Acts of God"

/ Priceless

What I still don't understand is the speed from my computer to the caching server over my lowly 3 Mbps second ADSL line, and for that matter over my wireless LAN.

It could be your local (internal browser) cache which would get speeds around that. My understanding the speed tests should generate a random data set for the download. If the randomizer for some reason didn't work at that time then the comparison of the internal cache to the data would match up and the browser would use the local cache. In other words, caching works by looking for changes from data coming from the site and the cache. If there is a difference, the new data will be downloaded and sent to the browser and the cache updated.

Oh well, I guess that I'll file it under "Random Acts of God"

That works also.