Internet Speed Tests - Results, comparisons, graphs, etc

tankman1989

Active Member
Reaction score
5
I thought I would start a thread where you can post the results (of any kind) of speed tests done on Speedtest.net. I only use that site b/c it offers so many awesome features with the account option to track results over time, multiple locations, etc. As you will see, the distance from my location to the server is 2,450 miles, which ensures that these servers are "out of network" which is a big deal when testing the TRUE speed of Internet connection (and why it's important). I can give analogies if anyone wants one, LMK...

If you have more than one ISP or even different service levels within the same ISP and even business vs residential, all that is interesting to see the difference, IMO. It can help a business owner decide which if the services will best suite his clients needs based upon real life numbers vs the numbers you are being told by the marketing dept. I'm getting 3.61Mbps download when I should be getting 24Mbps down (which is what they sold me but after about 20 hours of tech installs and problems, had to drop me to 12/2 and now I require another CONSISTENT ISP offering 12/2 to get close to the 24 that I needed.). DON"T TRUST THE SALES REPS UNTIL YOU DO YOUR OWN RESEARCH!!!

I would HIGHLY advise that you create an account (they are all free!!) with Speedtest.net and login when you do your testing. There are a lot of really nice features available to members such as tracking all previous tests by time, date, IP address, ISP, testing server, dl & ul speeds, ping/latency, distance, etc. All of this can be used to create all kinds of graphs or averages depending upon what you decide to use for variables. It's really great. You can also compare against other members (of same service or different services) to see how you stack up. MOST IMPORTANT OF ALL is that you can use it as a means to get credit for your service if the quality does not meet your needs. With this you have PROOF of their network quality which is why I tell you to pick a server OUTSIDE of your ISP's network (might take some digging, but ask and we can help). It can even get you out of a contract for failure to provide on their end.

This would be helpful for your clients if they are having problems with service & there is an alternative. Have them run random tests which you later have them show the ISP. This is what I'm doing now. Then, sign the customer up with the competitor (one you TRUST and KNOW) and get the referral fee - anywhere from $100-$1200 depending upon the area & length of contract!

When you post your results, it would be great if you post the speed package that you have, the service, provider and the town/city count and state if you wouldn't mind doing so. It gives a better representation of where these services are being used.


AREA - LANCASTER PA

Penn Tele Data 12/2Mbps CABLE package


Windstream 12/2Mbps DSL Package


Penn Tele Data 12/2Mbps CABLE package


Windstream 12/2Mbps DSL Package


*****Penn Tele Data 12/2Mbps CABLE package
*****100mil from home - in network but was SLOWER!!!???
*****
2594822075.png
 
Last edited:
Based in Bradford, West Yorkshire, United Kingdom

BT Infinity 2 - Exchange to Street FibreOptic - Street to House DSL over Copper (Upto 76Mb up / Upto 19Mb down)


Home BT Infinity - Server 17miles
2594955561.png


Server Europe
2594928532.png


USA Server

2594932126.png
 
Last edited:
To all the above posters: I added this to my OP:
When you post your results, it would be great if you post the speed package that you have, the service, provider and the town/city count and state if you wouldn't mind doing so. It gives a better representation of where these services are being used.

Put in what you like and leave out stuff if you feel you should. DSLreports openly lists all of this and more (Ip addresses) so none of this should really matter.
 


I'm located in Richmond, VA and have Verizon FiOS Quantum - Promised 50/20 and definitely getting it. :cool:

:(My lower lip is quivvering after seeing that. Me not so happy. Me Jealous. Me even more mad sales 8!7@# lied about giving me 24/4 and only getting 3/1.:mad: Just a little bit of difference
 
Not sure what your are getting at. Is sarcasm meter not working with 18 hours of no sleep or just getting dumber:p

What i'm saying is although speedtest.net has improved dramatically over the years in terms of multithreaded speed tests it is still not the best way to measure maximum speeds.....though I admit while testing it right now I see no flaw...might be because its not during peak hours.
 
What i'm saying is although speedtest.net has improved dramatically over the years in terms of multithreaded speed tests it is still not the best way to measure maximum speeds.....though I admit while testing it right now I see no flaw...might be because its not during peak hours.

Thanks for the info! Keep us updated on this type of information and if it is OK with you, I will add it to the OP as it arises, so first time readers get the "jist" of the thread w/o having to read it all, just read the first posting. Is that Ok with you?

The other option was to test with Speakeasy which I ruled out for MANY reasons. It just wasn't flexible enough, didn't keep track of cumulative tests and the server selection was pretty slim making it difficult to find one outside of a network at times. With only 8 servers, the ISP's can easily prioritize traffic to those servers to ensure fast testing times, but with Speedtest.net, there are so many, that change from time to time, that it is much more difficult.

Is there a better testing method which you could recommend in addition or as a replacement to this test?
 
Thanks for the info! Keep us updated on this type of information and if it is OK with you, I will add it to the OP as it arises, so first time readers get the "jist" of the thread w/o having to read it all, just read the first posting. Is that Ok with you?

The other option was to test with Speakeasy which I ruled out for MANY reasons. It just wasn't flexible enough, didn't keep track of cumulative tests and the server selection was pretty slim making it difficult to find one outside of a network at times. With only 8 servers, the ISP's can easily prioritize traffic to those servers to ensure fast testing times, but with Speedtest.net, there are so many, that change from time to time, that it is much more difficult.

Is there a better testing method which you could recommend in addition or as a replacement to this test?

For download speed? I always just test with tons of threads on newsgroup servers. If I can't reach the max download speed using that then I know i'm not getting the advertised speed for certain.
 
I don't put much value in speedtests other than giving a brief impression of what your performance "might be at some given moment".

They are incredibly inaccurate for "bursting" cable packages....when I had Comcast a few years ago I would run speedtests in the 70 or 80 and even 90 meg range. That was with their 8 meg package. "Powerboost" for the first 10 megs of a download.

I'm on wireless right now, a floor up in my house, and one of my kids right now is torrenting and playing online I'm sure. Plus several other computers plus tablets plus smartphones on. I have plugged right into my router in the past with nobody else on and saw all 18 megs.

AT&T U-Verse 18 meg package
 
I don't put much value in speedtests other than giving a brief impression of what your performance "might be at some given moment".

They are incredibly inaccurate for "bursting" cable packages....when I had Comcast a few years ago I would run speedtests in the 70 or 80 and even 90 meg range. That was with their 8 meg package. "Powerboost" for the first 10 megs of a download.

An excellent point! I forgot all about powerboost.
 
An excellent point! I forgot all about powerboost.

Before Comcast even announced "PowerBoost"....I had heard about it from friends that worked for them. I was in a neighborhood that was one of the first areas to receive that feature. They totally opened the throttles, there was no "choke"...like they added later. So on my measly 8 meg cable package, yeah I was getting darned close to 100 megs. At that time I was into IPCop and Endian for my router at home...it was running on a P3 1.0 or something like that. No home grade Stinksys or Nutgear router to be a bottleneck.

I recall having a picture of 91 megs or something...I looked at my FTP pictures directory up at Speedguide.net and all I can find is one at 74 megs.
But those with observant eyes can tell the time era of the picture by seeing the Eset NOD32 2.7 icon in my systray, and at that time I was using Opera browser.

This is 74 megs on an 8 meg Comcast back in ..oh I forget the year.
Those familiar with cable "bursting" packages...of course once you've download some set amount (back during this time it was 10 megs)...your download speeds would throttle down to what your package is rated for (in my case, 8 megs). They later put some throttling on it so the "boost" speeds would peak at around 25 megs.
powerboost2.JPG


Now here's a picture of a Comcast business grade account, Comcasts Docsis 3 50/10 package. 50 megs down, 10 megs up. Here is a screenshot of my laptop downloading a 5 gig ISO from Microsoft...I am over 1 gig into the download, so obviously now "powerboost" coming into play here. It is on a network of about 60 computers, and I have QoS on their Untangle firewall.

You can see the download rate at just over 6 MB/s.
The math equals 48 Mb/s...pretty darned close to the 50 meg package. And considering the load on that business network happening at the same time, not bad.

hospicecomcastdown2m.jpg
 
At home (Elkhart, IN) on my netbook... connected via WIFI through a WIFI repeater


(here's Comcast's internal speed test)
231527872.png



At home on my server... hardwired to router

231528282.png


Phone gets about 10 down 6 up connected via wifi to the main router


Not sure if the Speedtest.com tests and the Xfinity tests are going to the same server or not, honestly.
Our plan is the Blast internet which just got upgraded to 50M down... but our modem is only DOCSIS 2.0 and doesn't support the increase. :(
 
Last edited:
Back
Top