<Didn't someone comment there was little demand for bandwidth????>
Reading from last back, I probably have time only for this response.
The article is a good description of current bandwidth limitations facing the net. Other than the "field of dreams" comment, which will EVENTUALLY be true, I thought it was well represented.
Now, let's consider (again) the difference between BACKBONE and local loop bandwidth.
Backbone bandwidth is needed due to the ever growing number of internet users. 20 million people sending email creates a major backbone bandwidth bottleneck. But email is NOT a bandwidth intensive application on the local loop (for a local user). See the difference?
Just because you see the term bandwidth shortage in print, it does NOT imply a need for high bandwidth for the local loop. In fact, higher local loop bandwidth will compound the backbone problem, and is a major reason why it doesn't do any good to increase local bandwidth - because you won't see it very much in real use internet throughput.
The question (which you never did answer) was how many minutes would having a 1Mbps modem save *you* in your 12+ hours on the internet each day? How many minutes would a 6Mbps save you?
Not sure how to answer this? This may highlight WHY this point is not clearer.
Let's think it through: Let's say you want to download a 6MB file, with a 28.8 modem (and we'll assume a RARELY achievable "ideal" condition of NO backbone delays), how long would this take?
First, 6MB is equivalent to 48Mbits. So with a 28.8 k(bits)ps modem,
this would take 1666 seconds or almost 3 hours. With a 1Mbps modem (and ideal conditions), this would take 48 minutes. With a 6Mbps this would take 8 minutes.
Pretty good, huh? Except you will rarely get this kind of throughput, and the backbone lag is more apparent with a faster local connect. In other words, the 28.8 may actually take 4 hours, the 1Mbps may take 2 hours and the 6Mbps may take more than 30+ minutes.
So the incremental increase in performance does NOT increase linearly with local access speed. Still worth it, you say? To some maybe. But how much is this REALLY worth? Here's where the REAL test comes in. When was the last time, you, an internet intensive user, downloaded a 6MB file?
I work in a data intensive field, on dedicated E1s transporting brain image data across the country, and I RARELY transfer files this large.
In my NON-technical use of the net, I almost never download files this large, and I am on the net *many* hours each day also. The largest file I downloaded was Netscape 3.3, and I went out for the afternoon and let my 28.8 chug along - no problem.
Do you see the point? So how often will Joe-1hour/day-internet-user need anything more than a 28.8K or a 53K?
Would it be nice to load that graphics intensive 50K page in a tenth of a second rather than in 2-4 seconds? Sure, but even THIS size file is not that common, and I wouldn't pay the extra $40-$130 more a month (or more) to be able to do so, in my Joe-5to10hours/week internet use (let alone the Jane5-10hours/month use). At least not in the next few years, when there is really no compelling NEED (ie., "killer app") to.
See my point?
Steve |
