What does my download speed vary so much from server to server?

[RBinVA]

Hello,
I just had installed cable modem with Charter, on a 30Mbps speed plan. Charter provided a new Arris Docsis 3.0 modem (combined for internet and telephone - MTA), and I have a Linksys Valet M20 wireless router (10/100, I believe) and a Linsys 5 port (unmanaged, 10/100) switch (with CAT6 cable) to connect to computers and HDTV in other rooms. When I check Download speeds on speedtest.net, my results will vary from 27-29 Mbps, wiith ping at 23-25 ms (actually hit 40 one time) to 12.5-13, with ping typicaly 42-46ms. Multiple tests were run over about 3 hours in the eveniing. The only difference I see is the server I connect to for the speedtest - I am in Suffolk, VA, and when I connect to the Reston, Va servers, I consistently hit the high DL speeds on all my computers (wireless and ethernet) around the house, whether I have 1 or all computers on line at the same time. {Results were the same whether on Firefox or IE and whether XP or Win7 professional.} But when the server is Richmond (Comcast), which is much closer, or Wash DC (same distance as Reston), I consistently hit 12-13 Mbps. Upload speeds are consistently between 3.0-3.19 for all tests. My question: Is it just the server that dictates the speed I am showing or could there be a bottleneck in my home equipment, or perhaps a difference in how my equipment connects to different servers? If home equipment is the likely bottleneck, please suggest some alternative software, tweaking and/or equipment changes.
Thanks

 

[eibgrad]

The server matters in several respects. First, you never know the loading of any server at any given time (there might be many other users using the same server, while the other is nearly idle). Secondly, you can never be assured the routing used to any server. That can change over time as well. The only reason it's generally recommended that you use a server that's physically closer is the assumption that it will require less routing (hops). But that's only a generalization. It's not necessarily true, or true all the time. The beauty of the internet is its flexibility to route/reroute as necessary. But that comes at the price of predictability when using it.

That's why it's so hard to assess your home network's performance by using only the internet. The internet is far less predictable, w/ too many variables and circumstances beyond your control. You can only assess your home network's performance LOCALLY (machine to machine), convince yourself that it’s sound, then see how it performs under varying internet conditions.

 

[RBinVA]

Eibgrad
Thanks for this reply - that makes sense not to depend on the internet to assess my equipment. I guess my real need is to determine what I need to do to be consistently interfacing with the internet at my maximum speed on my end - knowing that I cannot choose the server I connect to. It is important because sometimes the speed is quite slow and interferes with tele-commuting work from home. (Was much worse with Verizon as the max DL speed available was 3Mbps). So the next questions would be: 1) how do I assess my LOCAL network performance - what should I be looking for, and 2) is it possible that my wireless router can affect what speeds I am connecting to the internet at, that is, would a newer gigabit router impact my interface with the internet differently than the current one I am using and allow for faster speeds? Or, is the router not even a factor, and the connection/DL speed is simply a function of the server and it's loading. If these questions are already commonly answered in another thread, I am happy to be referred to those links. Again thanks for your early response to my questions. [ Note - I was not sure my question had posted earlier today, and I ended up posting it again tonight- the second post can be deleted by whomever controls this site.]