Internet Speed

[Shoe]

I have a wireless router with hard-wired desktops and a wireless
laptop and printer. I ran speed checks from several different web
sites yesterday and they were all pretty much in agreement. Download
speed was terrific, around 18,000 Kbps. Upload speed was absolutely
awful at 350 Kbps. I got similar results on the wireless laptop. Any
thoughts on cause of slow upload speed?

 

[Peter Taylor]

I have a wireless router with hard-wired desktops and a wireless
laptop and printer. I ran speed checks from several different web
sites yesterday and they were all pretty much in agreement. Download
speed was terrific, around 18,000 Kbps. Upload speed was absolutely
awful at 350 Kbps. I got similar results on the wireless laptop. Any
thoughts on cause of slow upload speed?

Did you try speedtest.net?

 

[Seth]

I have a wireless router with hard-wired desktops and a wireless
laptop and printer. I ran speed checks from several different web
sites yesterday and they were all pretty much in agreement. Download
speed was terrific, around 18,000 Kbps. Upload speed was absolutely
awful at 350 Kbps. I got similar results on the wireless laptop. Any
thoughts on cause of slow upload speed?

Might be your ISP. Most of them of much lower upload speed vs. download
speed.

If all your machines are behaving the same, I would suspect the problem is
on your end. Last test to try to to take the router out of the equation and
hook one of your machines directly to the ISPs modem.

 

[Tim Slattery]

I have a wireless router with hard-wired desktops and a wireless
laptop and printer. I ran speed checks from several different web
sites yesterday and they were all pretty much in agreement. Download
speed was terrific, around 18,000 Kbps. Upload speed was absolutely
awful at 350 Kbps. I got similar results on the wireless laptop. Any
thoughts on cause of slow upload speed?

Many ISPs operate asymmetrically, which is exactly what you're
describing. The idea, I think, is to prevent you from running a web
server from your home computer, which could eat up way too much of
their bandwidth.

That's just speculation, but if there's a better explanation for that
policy, I'd love to hear it.

 

[Mike]

I have a wireless router with hard-wired desktops and a wireless
laptop and printer. I ran speed checks from several different web
sites yesterday and they were all pretty much in agreement. Download
speed was terrific, around 18,000 Kbps. Upload speed was absolutely
awful at 350 Kbps. I got similar results on the wireless laptop. Any
thoughts on cause of slow upload speed?

This has what to do with Win 7? Complain to your ISP?

 

[Char Jackson]

Many ISPs operate asymmetrically, which is exactly what you're
describing. The idea, I think, is to prevent you from running a web
server from your home computer, which could eat up way too much of
their bandwidth.

That's just speculation, but if there's a better explanation for that
policy, I'd love to hear it.

I believe you have it backwards. It's not that they made the upload
speed small in order to stop you from running a server, but instead
some ISPs don't want you running a server because the upload speed is
small.

It's a simple matter of economics. When the ISPs, the cable companies
and DSL providers, for example, were getting ready to design their
systems, they analyzed the expected user behavior and determined that
most residential users would download far more data than they would
upload, so the download pipes were built bigger than the upload pipes,
saving the ISPs a lot of money in the process. Having done that, it
became necessary to limit the user's ability to upload large amounts
of data because the pipes can't handle it.

That's slowly changing now, with upload speeds being given a little
more room to breathe, but it's still wildly asymmetric, and rightly
so. In my area, the business class accounts are symmetrical, which
also makes sense.

 

[Ken Blake]

Many ISPs operate asymmetrically, which is exactly what you're
describing.

Yes, asymmetrically, but nowhere near as much as 57 to 1, which is
what he described.

My ISP is Comcast. I just ran a test on speedtest.net. I got 18.45
Mb/S DL and 4.39Mb/S UL. That's a ratio of only 4 to 1.

Still, I would recommend that he contact his ISP, and ask them why
that's what he's experiencing. It can't hurt, and they might have a
clue for him.

The idea, I think, is to prevent you from running a web
server from your home computer, which could eat up way too much of
their bandwidth.

That's just speculation, but if there's a better explanation for that
policy, I'd love to hear it.

I don't know either, but your speculation makes sense to me.

 

[Dave \Crash\ Dummy]

Many ISPs operate asymmetrically, which is exactly what you're
describing. The idea, I think, is to prevent you from running a web
server from your home computer, which could eat up way too much of
their bandwidth.

That's just speculation, but if there's a better explanation for that
policy, I'd love to hear it.

Most ISP's operate asymmetrically because most home users operate
asymmetrically, downloading much, much more than they upload.
Discouraging servers is just an incidental benefit. (My 2 Mb upload speed
doesn't discourage me all that much.

But I agree with Ken. The asymmetry is usually more on the order of 10:1, or
less, not the imbalance Shoe reports.

 

[SC Tom]

Yes, asymmetrically, but nowhere near as much as 57 to 1, which is
what he described.

My ISP is Comcast. I just ran a test on speedtest.net. I got 18.45
Mb/S DL and 4.39Mb/S UL. That's a ratio of only 4 to 1.

Still, I would recommend that he contact his ISP, and ask them why
that's what he's experiencing. It can't hurt, and they might have a
clue for him.




I don't know either, but your speculation makes sense to me.

Depending on who I use to test, my results vary wildly. I can get as good as
6:1 and as poor as 183:1. My advertised rate is 5Mb/S DL, and .5Mb/S UL.

Charter (my ISP) gives me 11.3Mb/S DL/1.0Mb/S UL.
Argonne Nat'l. Labs 6.04/1.05
Speakeasy.net to ATL 10.31/.1
to DC 14.68/.08
to NYC 11.73/.06
Speedtest.net 10.41/.14

And that's during the day which gives me my best times. If I do it late
evening, I sometimes get speeds so slow it locks up IE.

Plus I love that my ISP shows my speeds as twice as fast as what I'm paying
for.

Of course, contacting my ISP is just about the biggest lesson in frustration
I have ever experienced in all my years of working with computers or the
retail public

 

[Dave \Crash\ Dummy]

Depending on who I use to test, my results vary wildly. I can get as
good as 6:1 and as poor as 183:1. My advertised rate is 5Mb/S DL, and
.5Mb/S UL.

Charter (my ISP) gives me 11.3Mb/S DL/1.0Mb/S UL. Argonne Nat'l. Labs
6.04/1.05 Speakeasy.net to ATL 10.31/.1 to DC 14.68/.08 to NYC
11.73/.06 Speedtest.net 10.41/.14

And that's during the day which gives me my best times. If I do it
late evening, I sometimes get speeds so slow it locks up IE.

Plus I love that my ISP shows my speeds as twice as fast as what I'm
paying for.

Of course, contacting my ISP is just about the biggest lesson in
frustration I have ever experienced in all my years of working with
computers or the retail public

I get the best and most consistent results with a mini speedtest.net engine
installed on my own website. My website is hosted by 1 and 1, but I have no
idea where it is located geographically. I installed it primarily to
test my website availability, but it works both ways. It has the
advantage of being completely free of any extraneous crap, like ads.
It's not really intended for public use, but you are welcome to it as
long as it is up.
http://thedatalist.com/speedtest/

 

[SC Tom]

I get the best and most consistent results with a mini speedtest.net
engine
installed on my own website. My website is hosted by 1 and 1, but I have
no
idea where it is located geographically. I installed it primarily to
test my website availability, but it works both ways. It has the
advantage of being completely free of any extraneous crap, like ads.
It's not really intended for public use, but you are welcome to it as long
as it is up.
http://thedatalist.com/speedtest/

Thanks! I tried your link and got 1218kbps/31kbps.

 

[Ken Blake]

Thanks! I tried your link and got 1218kbps/31kbps.

Typo? Did you mean Mbps, not Kbps?

I tried it too. I get 11902 Kbps and 331 Kbps. Our numbers are very
close if you meant Mbps.

 

[David]

There are technical reasons for asymmetric upstream and
downstream rates. For cable, the upstream bandwidth on the
physical cable line is much greater than the downstream.
This is fixed and determined by the cable facilities and the
amplifiers that power them so whatever portion of the
upstream bandwidth (shared by the TV services) the company
allocates to data access can be quite large. For DSL, the
bandwidth of the line has a certain up + down capacity and
the company decides how much goes up and down. This ratio is
time consuming to change dynamically (because of training)
so it is usually fixed for a certain customer to a smaller
upload rate and a greater download rate.

David
..

 

[Dave \Crash\ Dummy]

Typo? Did you mean Mbps, not Kbps?

I tried it too. I get 11902 Kbps and 331 Kbps. Our numbers are very
close if you meant Mbps.

As I said, I don't know where the server is, but I'm in the northwest
corner of North Carolina, and I get 20487 kbs and 2129 kbs, which is
what I am paying for. I rarely get readings like that at any of the
popular speed test sites.

 

[Dave \Crash\ Dummy]

As I said, I don't know where the server is, but I'm in the northwest
corner of North Carolina, and I get 20487 kbs and 2129 kbs, which is
what I am paying for. I rarely get readings like that at any of the
popular speed test sites.

As close as I can tell, my website host server is in Atlanta, Georgia,
which would be favorable for me.

 

[SC Tom]

Typo? Did you mean Mbps, not Kbps?

I tried it too. I get 11902 Kbps and 331 Kbps. Our numbers are very
close if you meant Mbps.

Nope, no typo :-(
I just reran it and got 1329/37 Kbps, so it's gotten better But when I
click on the link to run it from Columbia, SC, I get 9.01/.16 Mbps. I'm
located in the SE tip of SC, near Parris Island (Ooo-rah!!!)

 

[Dave \Crash\ Dummy]

Nope, no typo :-( I just reran it and got 1329/37 Kbps, so it's
gotten better But when I click on the link to run it from
Columbia, SC, I get 9.01/.16 Mbps. I'm located in the SE tip of SC,
near Parris Island (Ooo-rah!!!)

Well, those suggested links are the three test sites closest to you.
When I run the test at the Columbia, SC site, I get 28.57 Mbs / 2.14
Mbs. I'd have to say your service sucks, semper fi, or not.

 

[KCB]

I get the best and most consistent results with a mini speedtest.net
engine
installed on my own website. My website is hosted by 1 and 1, but I have
no
idea where it is located geographically. I installed it primarily to
test my website availability, but it works both ways. It has the
advantage of being completely free of any extraneous crap, like ads.
It's not really intended for public use, but you are welcome to it as long
as it is up.
http://thedatalist.com/speedtest/

Speakeasy.net has a good speed test:
http://www.speakeasy.net/speedtest/

Using the New York server gives me about 25 Mbps down, and 1.9 Mbps up.
Chicago and DC are around 20 down, 1.9 up. I'm near Pittsburgh.

 

[SC Tom]

Well, those suggested links are the three test sites closest to you.
When I run the test at the Columbia, SC site, I get 28.57 Mbs / 2.14
Mbs. I'd have to say your service sucks, semper fi, or not.
--

I've known that for quite a while, but unfortunately, they're the only game
in this part of town. Although I guess if I'm paying for 5Mb and I get
better than 6 or 7 on most test sites, I can't complain too much. And since
I don't do too much uploading, it doesn't matter much that its speed sucks
the big one big time.

 

[Char Jackson]

There are technical reasons for asymmetric upstream and
downstream rates. For cable, the upstream bandwidth on the
physical cable line is much greater than the downstream.

Other way around - downstream bandwidth is much greater than upstream
bandwidth.