Bandwidth and consequently dedicated server prices have been plummeting every day primarily because of Cogent

If this article is true, will it mean that bandwidth prices and overall web hosting prices will become reasonable?
http://www.americasnetwork.com/americasnetwork/article/articleDetail.jsp?id=21161

Cheers
Balaji

I really don't think Cogent is going to go under. Not anytime soon anyways.

What I could see however... may be Cogent working on getting most of these lines setup, and these clients on board, and then raising their prices.

So you may be right, prices may go up yet, but I think it will be a long time before that happens, and chances are somebody else will try to offer the lowest cost fiber connectivity before too long.

Just my 2 bits.

i see more of the same

cheap bandwidth
newbies
burn outs

Originally posted by MotleyFool
overall web hosting prices will become reasonable?


Web hosting companies all but give their hosting away, how much cheaper can it get?

my goodness!

I was telling we need to see higher prices to make a decent living

Cheers
Balaji

I dont understand why your complaining! It's like you want to pay more! It's simple economics: The more capacity bandwidth providers have the cheaper they will sell bandwidth for! Providers spend the late 90's expanding their data pipes because they though there would be huge demand for them because of the internet explosion. Now thousands of miles of "dark fiber" lay there so they have to do something with it. I'm not sure why you didnt expect this, I think in 10 years we will be buying a terabyte for $1 not a gigabyte. It might sound crazy now but it's not. As routers get faster and faster and more fiber gets laid bandwidth will get cheaper and cheaper.

I sort of know where your coming from though as I myself am starting a hosting company I understand that competitors are selling hosting for dirt cheap, and you know what? I dont care! Because I know that in 6 months they will be gone and i'm still going to be here. :)

Originally posted by TheException
I really don't think Cogent is going to go under. Not anytime soon anyways.

Hmnn... I never, ever, assume anything about the financial state of companies - especially those who have big capital investments and uncertain cash flow.

http://bigcharts.marketwatch.com/quickchart/quickchart.asp?symb=coi&sid=0&o_symb=coi

Enron and all the other blow ups just go to show you that cooking the books is not a rare event. So I trust charts more than financials and certainly more that gut feel.

Could turn into
http://bigcharts.marketwatch.com/quickchart/quickchart.asp?symb=WCGRQ&sid=157372&time=

Level 3 looking poor, a failed test of top and poised to test the recent significant lows and watch out.

http://bigcharts.marketwatch.com/quickchart/quickchart.asp?symb=LVLT&sid=102815&time=

Genuity, yikes.
http://bigcharts.marketwatch.com/quickchart/quickchart.asp?symb=GENUD&sid=1152908&time=

Clearly there is more bandwidth than costs and revenue justify. That can't go on forever.

Originally posted by mlip129
I dont understand why your complaining! It's like you want to pay more!

Yes, indeed that's what I would want. It sounds funny but is also simple. Me as a dedicated server provider I can say that I make more money (in relation to quantity) form 270$/mbit Yipes bw then 30$/mbit Cogent. Profits are percentages of the costs and therefore the best thing that could happen to me is that Cogent went Chapter 11 and prives returned to an 200$/mbit average... then it would be profitable again.

As an out of work telecom guy lemme let you in on a few things...

ALL transport is overbuilt!!!

AT&T has rings upon rings upon rings and in each ring are HUNDREDS of little pieces of glass called dark fiber...why is it dark? cause they can't sell it!

Why can't they sell it? because they are now using spectrum multiplexing to make a single strand carry not an OC3, not an OC48 or even an OC192. No, now they can put (give or take working from memory) 48 seperate beams, on a single fiber. That's so much bandwidth that it'll be YEARS before they can even come up with ways to use it all....

Telcom's biggest problem is they have to protect old buisiness. If they want to charge you a grand a month for a T1 for phone lines, they have to charge 15k or more for a T3. If they charge that much for a T3, they can't verywell sell you an OC192 for anything even within reason.

My point?

It'll be a long time before the bandwidth glut is over.

Will Cogent make it to the end of the glut? Maybe, Maybe not....

Originally posted by Shin
As an out of work telecom guy lemme let you in on a few things...

ALL transport is overbuilt!!!

AT&T has rings upon rings upon rings and in each ring are HUNDREDS of little pieces of glass called dark fiber...why is it dark? cause they can't sell it!

Why can't they sell it? because they are now using spectrum multiplexing to make a single strand carry not an OC3, not an OC48 or even an OC192. No, now they can put (give or take working from memory) 48 seperate beams, on a single fiber. That's so much bandwidth that it'll be YEARS before they can even come up with ways to use it all....

Telcom's biggest problem is they have to protect old buisiness. If they want to charge you a grand a month for a T1 for phone lines, they have to charge 15k or more for a T3. If they charge that much for a T3, they can't verywell sell you an OC192 for anything even within reason.

My point?

It'll be a long time before the bandwidth glut is over.

Will Cognet make it to the end of the glut? Maybe, Maybe not....

Ok, but it wont be like that forever, a T1 wont cut it in 5 or 10 years and thats when prices will drop. Also I dont think providers will let dark fiber lay around forever. There will come a time where a average person will use a 100mbit line. I know that sounds insane but when tv on demand and similar high bandwidth services such as high quality video streaming come out people will need it. Think of how many people use 56k modems right now, you have to wait for everything but why? The capacity to for very high bandwidth is here! People will soon realize they need more, and dsl and cable wont cut it!

mlip129,

Do you know what the cable that the "cable company" brought into your house can do?

It can do at LEAST 45MBPS or = to a T3. (remember that before people wanted it phone lines could only do 300baud)

It'll be a LOOONNNNGGG time before we have fiber in our houses. It is just too expensive to lay the "last mile"

What you are saying is the exact reason that we are in the situation we are in....

All the companies had people like you saying the same things that you are so they built and built and built some more....

We were all supposed to be using a meg a second as I type this...But what am I gonna use it for?

I'm one person, even on 4 PCs, am i really gonna download mpegs, listen to internet radio, while playing an online game, talking on an internet phone, while watching CNN and checking my "online streaming quotes"?

I could really go on and on with this thread...in the article they talk about what basicly comes down to missed deadlines of getting the fiber to the actual customer...

I'd bet that 99.999% of that is not the fault of Cogent...

Maybe they are going by poles. Well the phone company or the power company owns those poles. So they need to get rights to use them. Maybe they need to dig, they need digsafes from the power company, the phone company, and the gas company.

Install dates are never firm in the telecom industry. They are missed every day by every company. There are too many variables for it to be any other way.

I am by no mean a networking expert, but I read an interesting article in Network Computing that while the bandwidth available is huge, the real bottleneck is the core of the internet connection, which is not too broad. They gave some examples on how a couple thousand people downloading movies could jammed the core of the internet. (Not sure about the number, can't remember)

Peter

Originally posted by Shin
mlip129,

Do you know what the cable that the "cable company" brought into your house can do?

It can do at LEAST 45MBPS or = to a T3. (remember that before people wanted it phone lines could only do 300baud)

It'll be a LOOONNNNGGG time before we have fiber in our houses. It is just too expensive to lay the "last mile"

What you are saying is the exact reason that we are in the situation we are in....

All the companies had people like you saying the same things that you are so they built and built and built some more....

We were all supposed to be using a meg a second as I type this...But what am I gonna use it for?

I'm one person, even on 4 PCs, am i really gonna download mpegs, listen to internet radio, while playing an online game, talking on an internet phone, while watching CNN and checking my "online streaming quotes"?

I could really go on and on with this thread...in the article they talk about what basicly comes down to missed deadlines of getting the fiber to the actual customer...

I'd bet that 99.999% of that is not the fault of Cogent...

Maybe they are going by poles. Well the phone company or the power company owns those poles. So they need to get rights to use them. Maybe they need to dig, they need digsafes from the power company, the phone company, and the gas company.

Install dates are never firm in the telecom industry. They are missed every day by every company. There are too many variables for it to be any other way.

I'm not saying that we need all this bandwidth now, but within the next 20 years we probably will. The problem is that the internet is built around dial up modems. People are using a 100k and under per webpage rule and thats why we still dont need it. But when the majority of people will have broadband things will change. When everyone has broadband there will be a demand for high quality video streaming and etc... I'm also not saying that every home needs a fiber connection, but the fiber will atleast be a good high speed backbone for when it needs to be used.

Originally posted by ckpeter
I am by no mean a networking expert, but I read an interesting article in Network Computing that while the bandwidth available is huge, the real bottleneck is the core of the internet connection, which is not too broad. They gave some examples on how a couple thousand people downloading movies could jammed the core of the internet. (Not sure about the number, can't remember)

Peter

This is actually a good point. The Internet is getting more and more meshed into a big soup. Can any regular user really tell the difference between AT&T and Sprint or Level (3)? It seems everyday that there are less and less small peering points, and more large ones where everyone goes.


People are using a 100k and under per webpage rule and thats why we still dont need it. But when the majority of people will have broadband things will change.

This is like saying, once we do away with paper, books will be longer. Also, traffic for sites will get harder to come by. Your site on Teletubies might be popular now and drawing 100 GBs of transfer, but as soon as 10 more Teletubies sites appear..........

Originally posted by UmBillyCord


This is actually a good point. The Internet is getting more and more meshed into a big soup. Can any regular user really tell the difference between AT&T and Sprint or Level (3)? It seems everyday that there are less and less small peering points, and more large ones where everyone goes.




This is like saying, once we do away with paper, books will be longer. Also, traffic for sites will get harder to come by. Your site on Teletubies might be popular now and drawing 100 GBs of transfer, but as soon as 10 more Teletubies sites appear..........

I'm not talking about how long a webpage is, but the quality of images and movies, look at how crappy streaming media is, but why? Becase bandwidth is expensive and even if it wasn't people dont have the connections to view high quality video.

Originally posted by UmBillyCord
This is like saying, once we do away with paper, books will be longer.

I have disagree with you and agree with mlip129. Once the majority of people begin using high speed connections everyone will start designing their sites for them. I have already seen many sites the say "optimized for high speed internet access".

People will fail to realize that even though they have high speed internet access and they can load the page in less than a second that not everyone will have that luxury. There are probably still people connecting using Baud rated modems. :eek:

It is just the course of the internet. When you were using a 14.4K modem did you see alot of graphic intensive sites? Probably not, but now you do. ;)

Once the majority of people begin using high speed connections everyone will start designing their sites for them.

Why?? What is the benifit to the average Joe Blow? Do you think every person who has a web site is some over-the-top, sharks with lazer beam, type person? No. Most sites are for content. They don't need streaming. They don't need crazy crap on their web site. Do you think the majority will say, "hey, everyone has high speed access. Lets design a 1 GB - Flash, streaming intro with a full version of In-A-Gadda-Da-Vida playing"? Sure some will. So be it. However the huge percentage will not. Look at your subscriber base. How many actually use streaming or other high bandwidth use content? So many people get wrapped up in what they like and want, and use this train of though on the masses. Simple numbers prove this wrong.

Am I saying no one would use more bandwidth intensive contant? No. But I am saying there will ntoi be some huge shift to high bandwidth use content. Unless of course so new software or technology comes around that everyone must have and it is the host who can deliver.

What is called "high bandwidth content" today will be the "everyday use" content tomorrow - just look at the past years up to now.

Our expectations rise with advancement of technology and so does the normality with which we use this technology.

I personally think this is a good thing.;)

Our expectations rise with advancement of technology and so does the normality with which we use this technology.

I know. Just look at wheel. :)

I ain't saying we are not going to progress, but *I* don't think some huge content explosion will happen thanks to bigger pipes into our homes. Hell, look at CPUs. I find it hard to locate software or programs that my 550 GHz can't handle.

I find it hard to locate software or programs that my 550 GHz can't handle.

No wonder photoshop runs so slowly on my machine...I'm about 549 gigs behind the curve ;)

Originally posted by UmBillyCord
Why?? What is the benifit to the average Joe Blow? Do you think every person who has a web site is some over-the-top, sharks with lazer beam, type person? No....

(clipped)

So many people get wrapped up in what they like and want, and use this train of though on the masses. Simple numbers prove this wrong.


The problem is that the average Joe Blow may not realize that not everyone can connect to the internet as fast as him. I have a customer who I spoke to who thought that everyone had high speed internet access like him. It was only after I explained that not everyone is using current technology (high speed internet) that he decided to remove the 300K image on his home page and accept that not everyone likes/wants to wait for his image to load even though he thought it was worth it.

It is the average Joe Blow that will make large pipes almost a requirement to go online. Large corporations and people who hire decent web designers are the only ones who will be able to keep things to a minimum and not require large pipes.

Once everyone gets high speed internet access people will "forget" to make sure it works well on slower users; therefore making the users get frustrated and have upgrade to the latest high speed technology to have a pleasurable experience while online.

It's also an evolution type of situation; When you were on a 286 and you started to see what the people could do with the "new" pentium computers, did you have to have one? Did you wait until you realized that certain things would work on your system but the could on theirs; eventually giving in and purchasing the lastest system?

Everything that you utilize on the internet/computers have gone through an evolution type of situation, including bandwidth. I think that we are very far from ultimate goal of everything instantly, but we are slowly begining to accomplish it; and as we slowly accompish this task people will find a way to use it. Which then creates the loop called evolution; something great must come out to be greater than the current greatest thing.

Just my... err $200.00 worth

Taking a bit of U-Turn back to the first question in the thread... If this article is true, will it mean that bandwidth prices and overall web hosting prices will become reasonable? I believe the majority of prices are already "reasonable". Looking at the last part of the question however... "overall web hosting prices"... even if bandwidth became honestly FREE we still need to pay for the humans and the hardware etc... If site owners (and hosts) continue to push the prices down "because bandwidth is cheap/free" how do we plan to cover the costs of ... "everything else"?

This thread is a good read.. no question about that. It's interesting to read the opinions of a variety of interests. The issue of bandwidth reliability and costs is important as well. However, I think we need to be careful judging the "overall cost of hosting" by the specific cost of bandwidth. Remember that the majority of hosts already charge less for the bandwidth then they are paying (with the assumption that their users wont use it all). If their costs were to go down it might actually put them in the position of making a "real" (as opposed to the perceived) profit. $5/month for 100 accounts may seem like a profit to someone paying $200/month for their server and sticking the extra $300 in their pocket but as they grow and/or the Biz/Tax laws get a hold of them the "profit" can quickly turn into a loss. Long story short, as I noted above.. We need to make a clear separation between the "cost of bandwidth" and the "overall cost of hosting. Just my... err $200.00 worth I'm not as rich as xerocity.com but there's my nickel anyway :D

As a consumer and student, I enjoy the low prices and the extra profit I make as a result. I can, however, sympathize with those who would like to see prices go up.

The internet is becoming more and more bandwidth intensive. For the last 5 years I have been on the same 28.8 connection. Back in the day I could get by; there was no flash, or unsolicited (by me anyways) streaming media or hires images. Now surfing is an all day job.

Just look at the internet growth over the last 10 years: http://www.zakon.org/robert/internet/timeline/Count_Host.gif
http://www.zakon.org/robert/internet/timeline/
Think of what is to come. It's a hard one to predict, and as the internet becomes more and more a part of our lives for both work and fun.

To tell you the truth, I liked the internet better back in the good old days.

"There will come a time where a average person will use a 100mbit line. I know that sounds insane but when tv on demand and similar high bandwidth services such as high quality video streaming come out people will need it."

I think 100Mbs is a bit extreme but the truth is there. There are so many examples of the technological march we have: 640Kb RAM, ~6Billion IPs, etc. etc. 100Mbs is a long long way down the line [excuse the pun]. However 512Kb will become standard in Euope and the US in the next 5 years. The next progression is surely, "telephone" calls purely over fibre, multiple simaltaneous pay per view/live TV, Radio, 'Internet Downloads', music steros direct downloads etc. etc. I'm sure a lot more can be pushed on b/w as well as new things thought up. The growth will continue and continue for end-user bandwidth. Just as 2.0GHz processors wwere inconcievable in 1975 so is 100Mbs home feeds.

However sure there is loads and loads of dark fibre floating around but there are still one or two bottlenecks with ISPs, locally say for cable operators etc. etc. at the moment if we wanted too say use 50% of the dark fibre avaliable, bottlenecks elesewhere would prevent its efficient use.

JS

Originally posted by Deb
Taking a bit of U-Turn back to the first question in the thread... I believe the majority of prices are already "reasonable". Looking at the last part of the question however... "overall web hosting prices"... even if bandwidth became honestly FREE we still need to pay for the humans and the hardware etc... If site owners (and hosts) continue to push the prices down "because bandwidth is cheap/free" how do we plan to cover the costs of ... "everything else"?

This thread is a good read.. no question about that. It's interesting to read the opinions of a variety of interests. The issue of bandwidth reliability and costs is important as well. However, I think we need to be careful judging the "overall cost of hosting" by the specific cost of bandwidth. Remember that the majority of hosts already charge less for the bandwidth then they are paying (with the assumption that their users wont use it all). If their costs were to go down it might actually put them in the position of making a "real" (as opposed to the perceived) profit. $5/month for 100 accounts may seem like a profit to someone paying $200/month for their server and sticking the extra $300 in their pocket but as they grow and/or the Biz/Tax laws get a hold of them the "profit" can quickly turn into a loss. Long story short, as I noted above.. We need to make a clear separation between the "cost of bandwidth" and the "overall cost of hosting. I'm not as rich as xerocity.com but there's my nickel anyway :D


Deb,

I think I didn't word my first question well enough...

When I came to WHT a year back, badnwidth prices at 2$/gb were considered good and affordable.. and Rackshack's 99$ for 300GB offer was incredible

Now , thanks to Cogent, we have servers at 120$ and fairly good ones at that with 300-400gb /mo

I am paying 130$ for a PIII with 35gb/mo bandwidth and having enjoyed 100% uptime for 130 days I do know that there are NOC's there are NOC's.

But I am not able to compete on price with someone offering 250mb / 8 gb for 3$ or so.

So when I mean reasonable, I mean resonable for all the parties.. NOC, Data center, ded server provider, host, and customer [for an economy/industry to be stable all players must get a reasonable profit and the customer should get a decent value for money]

Price wars are a direct result of excessive supply over demand and it is happening just too much in WHT

Those who can hold their own will come on top , provide great service, good plans and reliable servers with near 100% uptime and make money

I am inclined to think that Cogent has contributed a lot to the glut in cheap servers [but I may be wrong ]


Well just my 2c !

Cheers
Balaji

The cheaper it goes, the more loosers and unserious peopl eare attracted... 13 year olds can pay ded servers from their pocket money. I'd really like the cheapest ded server to be like 500$/month or more so it becomes a BUSINESS again and not the joke it is now... :(

I did a troll on Google's Newsgroup DB a few months back. I came across some posts on the cost of "Internet" access in the mid-80's [contemporary posts] mainly concerned with getting data to the USSR.

The nominal costs of 1GB of Data Transfer I worked out was inexcess of US$500,000/GB and pushing towards US$1,000,000 when you take into account inflation. Is that serious enough:)? Of course they were only dealing in very small chucks of Data at the time 0.5Kb-5.0Kb etc.

Just goes to show that 100Mbs pipe may not be so far-sighted now for the end user:eek:.

JS

In all the discussions of how much should Internet bandwidth cost, i don't see one mention of the fact that bandwidth is more than just a piece of cable that runs between two locations, and then gives access to "the Internet".

Bandwidth is not a commodity, any more than web hosting is. It is a service. And as with anything else, some is better than others.

And unlike the sort of end-user Internet access that home users or businesses subscribe to, hosting facilities are purchasing Internet transit, which is more complex (and more expensive) than Internet access.

You are purchasing the right for your data to pass through their larger network on it's way to other networks. You are also contracting for them to run BGP with you.

Until their takeover by Worldcom and their service's resulting slide into averageness, UUNet used to be the ISP's ISP. They charged 3 times the going rate, and it was worth it, their service and their network was the best.

Some providers like Savvis and Allegiance (formerly Intermedia) have really sucky networks. Others cost the earth, but provide exceptional transit, their networks are either fast and well run, or they have a high percentage of hosts connected to them, or they have high-end SLA's.

We get some cheap transit bandwidth from a NAP in our area. No frills, no SLA's, just cheap bandwidth because they are a telco facility and we've got a rack of routers and such there.

Cheap bandwidth will always be available, good bandwidth will always cost more because it's the cost of the things on top of that raw bandwidth that add value and cost to the service.

The problem is that 98% of people won't pay extra for "better bandwidth"

99.9% of internet users can't tell the difference between "better bandwidth" and "cheap bandwidth"

While surfing at home, can you tell the difference? Do you click on a link and just KNOW that the site is using cheap bandwidth? Can you tell if it's the server pushing the images slowly, or the webmaster that didnt' optimize well enough, or if it's the bandwidth?

Can my mother on her AOL dialup really tell the difference between my Rackshack powered website and your dual p4 2 gig of ram scsi HD with expensive bandwidth site?

i agree

does the average user know if your on a super box or a catbox?

the only thing they do know, but only part of the time,

is if the noc your in has a crappy connection or not enough bandwidth and things get real slow or down all together

been there dun that too

we moved to another noc, and the whining stopped almost immediately....

Originally posted by Shin
The problem is that 98% of people won't pay extra for "better bandwidth"

99.9% of internet users can't tell the difference between "better bandwidth" and "cheap bandwidth"

While surfing at home, can you tell the difference? Do you click on a link and just KNOW that the site is using cheap bandwidth? Can you tell if it's the server pushing the images slowly, or the webmaster that didnt' optimize well enough, or if it's the bandwidth?

Can my mother on her AOL dialup really tell the difference between my Rackshack powered website and your dual p4 2 gig of ram scsi HD with expensive bandwidth site?

In that case.. Why bother using expensive bandwidth at all?!:confused:

There won't be money in bandwidth. Same in computers where chips, RAM and drives tend to 0.

Bandwidth tends to 0, as well.

There is money (big money indeed) in services: tuning, configuring, designing, administering, helping, installing, consulting, and services, services, services.

Regards.

People are now focussing on "bandwidth" as a terminology since people equate bandwidth as a commodity which is wrong, and misleading. It's the infrastructure that is the determining factor, and some people such as hardcore gamers(lan, internet, etc), are on the ball when it comes to this issue and understanding what is the determining factor nowadays of a server and where its at in the world connectivity roadmap.

Another factor is the route and intermittent bottlenecks that do occur each day, with people saying the host's servers suck or are down a lot when it's often a localised network bottleneck. All these factors attribute to perception, and so saying people dont care about infrastructure aka bandwidth is a tad unrealistic.