Quote:
Originally Posted by Kymmy
Totally different technologies. You can't really compare the differing media, distances, freqs
|
Well, that's a very high level reason for the difference and the sort of answer a politician or 'IT' manager might give. But what's the real technical explanation?
---------- Post added at 21:53 ---------- Previous post was at 21:19 ----------
Quote:
Originally Posted by MUD_Wizard
It will depend on the O/S you're using?
Windows 7 and upwards has TCP Window scaling. Google it.
Congestion on your local VM connection is also a factor.
If your repeated tests via http://www.thinkbroadband.com/speedtest.html never get a green line similar to the yellow line then you may have local congestion on your DOCSIS channels. Or it could be congestion over your wifi.
My connection over wifi:
This will also give you a clue about factors when connecting to web servers at a distance: http://bradhedlund.com/2008/12/19/ho...istance-links/
The computer and cable modem will need to ramp up the throughput by negotiation with the end server.
Most if not all servers are rate limited. That is they will only deliver a limited speed per connection to the server. So multiple connections are required in order to get more throughput. |
Thanks for the reply,
Well, I've used Win 7, 8.1 and a Debian OS, all with similar results.
Your comment about congestion is an interesting one. I've used the Thinkbroadband speed tests and at 6am get a graph that's very similar to yours. After that my '1x' speed is circa 30Mb/s when the '6x' stays at 150Mb/s - why (back to my original question)?
I get a very good Thinkbroadband ping graph; at all times; even when I can only get 20Mb/s download on ftp. For me that's very interesting.
Seems to me that VM significantly prioritise IP packets based on historical rates to particular TCP sockets.