Comcast 250GB Cap Goes Live October 1 → our competitive market

well, the march backwards starts Oct. 1.

caps beat the hell out of capex to add capacity don't they?

wonder how much Robert's bonus will be for the period beginning Oct. 1?

250GB is so bad. Right now its okay but not for long when HD Youtube and such goes live. Once inforced I dont think they will ever up it.

From Comcast's POV, caps beat customers downloading so many videos, they stop buying HBO (and other TV) from them.

It's a lot better than the 20GB cap Time Warner is "testing" on their Standard customers in Beaumont.

The real issue to take from a 250GB cap is not whether or not people are going to come close to hitting it in October 2008 when it gets implemented, because the overwhelming majority of customers won't.

The issue is what happens in 2009, 2010, and 2011. People aren't hitting 250GB a month using the majority of today's web applications because so few of them consume that amount of bandwidth. If you're running a torrent server or engaged in lots of online backups, you could easily hit this cap when it gets implemented. But the average person web browsing or watching YouTube will not... yet.

What will change all of this is streaming HD video content which is precisely where the next generation of applications are going to make an impact. It takes several gigs to deliver an hour of programming in HD, and as set top boxes connect to the net, consumers will easily consume many times the bandwidth they do right now.

Comcast's cap starts high and seems generous, but quickly becomes an impediment, particularly in establishing a precedent for usage caps, and likely will not grow with the times.

The abusers Comcast always likes to talk about are a tiny minority on Comcast's network, but the company punishes their entire customer base to deal with a few "problem customers." This, even after the company has a track record of dealing with users who consume "excessive" bandwidth.

So what is the real agenda here? To control "abuse" on their network which they were already doing, or attempt to manipulate the marketplace with caps, limits, and new revenue streams, particularly impacting the potential competition the video side of their business faces from a next generation Internet with high quality video streaming.

On one hand, Comcast has a new ad slogan opportunity ... "Well, we're not as bad as Time Warner" ...

on the other hand, Comcast never competes with Time Warner.

"It's a lot better than the 20GB cap Time Warner is "testing" on their Standard customers in Beaumont."

Maybe so, but that's like saying having your hand torn off is better than losing the whole arm...*sigh*

Right. They could have just made it 20GB and called it a day, instead they made it a reasonable 250GB.

Maybe their customers will stop using their service altogether and use Verizon/AT&T ... and they won't get any service from them.

A lot higher than Cox's 60 GB cap on Premier Tier, although they RARELY enforce it....

At this point in the day, we only have Karl's insider's information to go on. We don't know if upload is included, nor if the cap will increase regularly.

But my gut so far tells me that today this DL Cap is somewhere in the neighborhood of "fair enough." The break-even point could be about 150-350 GB (such a wide range because much of the cost is in infrastructure).

I still don't think customers want caps -- so Comcast has disappointed today. But if they insist on having them, disclosing them is the right thing to do.

I'll trade my 5gb frontier cap for the 250

Comcast usually will beat their speeds unless its FIOS.

Can't have your cake and eat it too Robb.

Actually, you're being overly generous. TW's proposed cap is the same as Frontiers--5GB.

That's an old saying, but what does it mean?

I can watch unlimited OnDemand movies. That takes bandwidth. If everyone on my block did it, each choosing different movies, we could very well recreate similar cost-of-service (upgrade) situations on the non-HSI that Comcast faces on the HSI side.

So I can have my cake and eat it too, no?

Here's the striking part. Bandwidth is becoming increasingly cheaper, while companies make it out to be more expensive. Sure, you got network costs (infrastructure, maintenance, etc), but better technology has still DRIVEN that down versus up.

If it were me, I think this would be a better method. THROTTLING.

Set a Cap, 250GB. Then throttle people's speeds after that to a point. SO someone has 10/1. If they exceed 250GB give them 2/1 instead for the remainder of the month. While not ideal, it slows down the user, freeing up "bandwidth" on the network for others. Seriously, ISPS need to upgrade to keep up with demand, but since that wont happen, THIS WOULD BE he next best thing.

I don't like this alternative either as I currently have 20/1.5 sold to me as unlimited service. why would I buy limited unlimited service? I do suppose it's better than just shutting down my modem but still...

Actually, your comparison is flawed for a couple reasons.

1. The major costs for an ISP are not always the raw bandwidth of their network, but the interconnect fees and uplinks with the rest of the internet.

2. VOD content is stored on a VOD cluster at your local headend. Internet data is...well.. on the internet. VOD goes from the cluster straight to RF to be broadcast on a VOD RF video channel. Internet data is a lot more shared.

3. Depending on the way your local node is setup, if everybody on your block was watching VOD at the same time, you could get an error because there is not another Video channel to put the new content on.

Data services and video services on a cable plant work a lot differently. Even on the backbone side of their network, a cableco may have a 100gig fiber link connecting all the systems within their area, but that link carries all their traffic, from video, to phone, to internet...and even their internal network traffic. The video feeds are going to be sent multicast to the different headends, so you are looking at 1 feed per channel for every system. Phone does take much space on their network. Internet however you are looking at....in theory... 1 stream per customer. The internet traffic is going to obviously be taking up a lot more traffic as a whole on their network than any other catagory of traffic when you look at it.

Even if you break it down per customer or per account, You use more bandwith on their network with your HSI, even if you just check your email and BBR several times a day, than you do for your TV service. (Figure even a HD channel running at 19mbps, multicast across their video backbone to every system in the area, and then given a dedicated RF video channel.... if you divide that bandwith across the 100,000+ customers in just your area, it's virtually nothing.)

Now I'm not going to say I'm for or against caps. I haven't honestly paid enough attention to the debate to be able to make a decision... but from the big network management side of things, I can see a need to have some sort of management practice in place in order to be able to manage the network....or even just to plan capacity upgrades. Obviously you can't lay down and turn up new fiber overnight, so I can see an advantage to having some manageable numbers available to help with the planning of capacity upgrades. (and 250gb is definately better than the much smaller numbers other ISP's have given us... and it's better in my opinion to know the limit than for them to keep the numbers hidden.)

Didn't I see someplace that that is kind of along the lines of what Comcast was proposing. If you exceeded the cap they would throttle you to "better than dsl speeds"?

Yah, I was thinking of the Frontier cap ...

It's all about the money, for corporations. Maximize profit, right?

You should be a member. We need informed people like you on this system. I could give you a "thumbs up" if you were a member. #1 is true for the TV side as well. The content comes from somewhere, and although we might have to make one step down away from the 3rd-part aggregation point in order to make my comparison, there are one-time and ongoing costs for interconnecting these things even within the same metro area.

#2 -- to RF, really? It's not sent to my box digitally and converted into RF there?

#3 and this is really the rough comparison I'm making -- both systems are (roughly) similarly subject to usage saturation. But we're making "hogs" out of the users who don't watch Comcast video products.

All in all, I agree with the gist of your message, though.

Part of Network Management is not selling more than you can reasonably expect to deliver. By putting in a cap, Comcast is saying that it can only deliver X. Now, they could rearrange things and avoid naming a cap, but doing so would expose Comcast to some competitive truths it would rather customers not know -- such as DSL may not be as fast second-by-second, but overall the Tortoise beats the Hare.

Cable VOD works by dynamically creating a QAM channel, typically unencrypted. You can sometimes find shows your neighbors are watching if you do a channel scan. Of course, if they pause the show to go to the bathroom, you're pausing too.

Yes, the only thing that can get to current generation digital set top boxes is RF over the coax.

VOD is just another channel on a QAM over RF. It will be combined with other channels, just as digital channel delivery occurs now over QAMs.

SDV is the same way, your set top box requests a channel and it is built on an IP Multicast network, but in the end, it is a channel riding on a QAM over RF to you.

Sometimes, depending on the cable company, if you have a TV that has a QAM tuner, and you auto-tune channels, you can pick up VOD streams that are currently playing while you were auto-tuning. If the cable company is not encrypting the feed to the cable box.

But running new fiber and adding new nodes to neighborhoods to fix the actual bandwidth choke points is a lot more expensive than buying more bandwidth from an backbone provider.

You can buy all the 10gbit or 40gbit circuits to the Internet you want, but that isn't going to give a node more than the 42mbit/sec that is available with DOCSIS 2.0.

So while you can probably get wholesale Internet bandwidth for $3-$5/mbit per month, that isn't going to help someone who is on a node with someone that uses their full 10-15mbit/sec 24x7x365.

No, I get that. I thought it was saying in #2 that the signal is sent as analog. (Mentally, I was interpreting the word RF as "analog signal.")

And so much for online backup services such as Dropbox, Mozy, iDisk, and whatever M$ calls their service.

Also be interested in how this changes telecommuting. This would be a step backwards if companies just tell you to come to work rather than paying for your internet.

The cost of physical connectivity is actually going up slightly, not down. The reason bandwidth is getting cheaper at the top is on that same physical run instead of just 1GigE you can run 10GigE or N*10GigE on the same physical fiber run if you use a WDM solution.

There is no upgrade in the DOCSIS world that gives a 1000+% bump in capacity, nor is the DOCSIS marketplace anywhere close to being as fiercely competitive as the Ethernet / SoNET / Carrier world.

The same factors that make bandwidth cheaper at the top unfortunately don't exist at the edge; at least not today.