Interesting article about Pulse8 and speedtests

http://www.ispreview.co.uk/index.php/2016/01/ookla-a...

Pulse8 dont have control over the network. Its Talk Talk if anyone who would be doing it

Speedtest.net not reliable anyway. Need a single thread speedtest for accurate results

Edited by bobble_bob (Wed 27-Jan-16 13:46:38)

If I use their speedtest its always slower then using the one on TBB.

Article doesnt really show much. Even they admit its not been tested with multiple users on multiple connections.

Unless more testing shows this happening, its abit of a non story

Pfffffft

Mulit-threaded tests will always show a better result than single threaded. As Ookla/Pulse8 use multi-threaded on their speedtest it is comparing apples to oranges...

Even my SamKnows box performs multi-threaded (they should consider single threaded as well) and it has been rock solid at 63Mbs for over a month (give or take a bit) which doesn't tell me much anyway.

All the article does is hurt Pulse8's hard earned reputation. My old Sky connection was no different to the article.

Use speedtest.net connection looked fantastic, use TBB it was absolutely awful. Based on TBB, TestMy and SamKnows reports I switched to Pulse8 haven't been happier, rock solid!

Im surprised how many people rely on speedtest.net and other multithread speedtests when they dont reflect a real world connection. All they so is saturate a commection and throw as much data down it to get a fast speed.

Fair enough not everyone is tech savy to know the difference but our own IT department wanted people to use speedtest.net to see if their connection was suitable for home working

And i think the article is unfair on Pulse8 when its ultimately TTB that control it

Edited by bobble_bob (Wed 27-Jan-16 15:12:44)

Inaccurate "articles" like this one really do my head in. Especially when I have had 20 hours sleep in the last 5 days!

An ISP with no control whatsoever over the network its subscribers use manipulating speed tests. That's one hell of a trick.

As far as accuracy being paramount to Ookla, hmm. A lot of people like Ookla tests as they give higher results than, say, ThinkBroadband. I prefer accurate over high for epenis enlargement.

Is it really that much of a shock if a speed tester hosted by the ISP and load balanced with local caches produces higher results than ones that are offnet?

PING www.supportal-test.co.uk (78.144.7.18) 56(84) bytes of data.
64 bytes from host-78-144-7-18.as13285.net (78.144.7.18): icmp_seq=1 ttl=60 time=6.48 ms
64 bytes from host-78-144-7-18.as13285.net (78.144.7.18): icmp_seq=2 ttl=60 time=6.58 ms
64 bytes from host-78-144-7-18.as13285.net (78.144.7.18): icmp_seq=3 ttl=60 time=6.62 ms
64 bytes from host-78-144-7-18.as13285.net (78.144.7.18): icmp_seq=4 ttl=60 time=6.70 ms
64 bytes from host-78-144-7-18.as13285.net (78.144.7.18): icmp_seq=5 ttl=60 time=6.49 ms

PING cs62.adn.xicdn.net (93.184.219.82) 56(84) bytes of data.
64 bytes from 93.184.219.82: icmp_seq=1 ttl=58 time=12.2 ms
64 bytes from 93.184.219.82: icmp_seq=2 ttl=58 time=12.0 ms
64 bytes from 93.184.219.82: icmp_seq=3 ttl=58 time=12.1 ms
64 bytes from 93.184.219.82: icmp_seq=4 ttl=58 time=11.8 ms
64 bytes from 93.184.219.82: icmp_seq=5 ttl=58 time=12.0 ms

Why don't you run a comparison?

http://www.myce.com/review/internet-providers-caught...

Just for arguments sake I did. Single threaded over ports 80 and 8080 came out at ~60mbps.

Ran the multi-threaded test over both ports 80 and 8080, again 60Mbps.