Will FiOS Eventually See Caps, Per-Byte Billing? → Why, they make money off those big tiers

Verizon doesn't cap because they don't need to by any stretch of the imagination. They have 4X the capacity of DOCSIS3 spread between only a tiny fraction of the homes that cable shares. Then there is just the total bandwidth the "hogs" consume. Again not a problem because Internet bandwidth is DIRT cheap, in the volume these ISPs use it's mere pennies per GB. It would take someone routinely blowing through 500-700GB/mo for Verizon to look twice at them. There aren't enough subs doing that to warrant a change in policy. IF one of those users were ever an issue, the current TOS/AUP impact the network clause would suffice. For seeders, there are other clauses like no serving that could be used.

So long as they don't raise speeds over much over 100Mb, their topology isn't suceptible to oversubscription. G-PON splits over 2Gbps between only 32 homes unlike cable which does 1/4 of that between hundreds and sometimes over 1000 homes.

Verizon's FTTH technology has the benefit of modern deployment so they will not need to resort to across the board capping. The bandwidth requirements of cable at the time of their deployment weren't but a fraction of what they are today so they're playing some catch up.

It seems to me that Verizon is going in the opposite direction...attracting hogs with pricy 15+ upload plans. So in a sense I guess you could say they're already doing a form of consumption billing by charging more to start. (in my market 15/15 is $65 and 30/15 is $140)

Actually GPON supports upto 64 subscribers. Just thought I would point that out since you enjoy talking about how much better FiOS is. Also 100Mbs would be over provisioning since the 2.488Gbs divided by 32 is only 77Mbs and 2.488Mbs divided by the 64 homes it's capable of is 38Mbs.

Again I don't argue that FiOS is the better technology but lets not mislead anyone into thinking FiOS is more than what it really is.

No, it's not 64 subscribers it's 64 homes and my understanding that Verizon is currently only deploying G-PON for 32 homes but even if it's 64 your argument is ludicrus.

Yes, if Verizon sells their service to 100% of the people who can get it and 100% of those people are using 100% of their connection all the time, they'll get 40-80Mbps instead of 100Mbps. Verizons current top tier is 50Mbps so by your numbers it is IMPOSSIBLE for Verizon to oversubscribe with their current plans.

Unless you think that Verizon can steal 100% of cable's HSI subscribers and converts 100% of their DSL subscribers, they won't have an oversubscription problem.

Use your head, cable is sharing way less than 1/4 of that bandwidth with hundreds possibly over 1000 homes and is just NOW having an oversubscription problem that is driving them to blanket capping.

You're still thinking at the neighborhood level when the over-subscription IS happening at the backbone level. The math just doesn't work otherwise.

Backbone bandwidth is scaleable and dirt cheap. Verizon has NO capacity problem nor anything even approaching a capacity problem.

Cable doesn't have a capacity problem outside the DOCSIS 1.1/2.0 upstream channel limitations but those would be remedied by DOCSIS3.

Make no mistake, cable's caps aren't about bandwidth, it's about killing video competition.

If "cable" was interested in controlling that upstream bandwidth (the seeding that is saturating the upstream channels) they would simply pull a Cablevision and throttle uploads for excessive users to 150kbps until those customers call. But they don't.

I don't buy that any of them have widespread downstream capacity issues. There will always be oversold pockets with cable (and their current topology) but I think that is the rare exception and not the rule.

Everyone acts as if these caps are necessary and they aren't. Upstream throttling with solve the seeding problem MSOs see.

Sure it's scalable. More fiber, bigger routers, more racks in the POP. More power, more AC, more batteries, bigger generators.

Dirt cheap? Not hardly when you're dealing at the macro level. Price out some carrier grade hardware (and don't forget some people to run them) and get back to us ok?

"Bigger routers"? LOL. More power? More AC? These devices are getting MORE energy efficient with every generation, not less. And how much more money per subscriber do you actually think it costs in electricity and air conditioning if they use 5GB/mo or 500GB/mo? Oh brother.

More fiber? They're already doing over 600Mbps PER STRAND with only 3 wavelengths. What makes you think they'll need more fiber? Racks in the POP? So what, you only need more racks if you have more customers, each paying over $50 whether they use their service daily or not.

The price of this regional hardware is MINISCULE compared to the revenues obtained with those devices over their lifespan.

Verizon has no capacity problem and nothing even approaching a capacity problem, even with 100% of their deployed customer base buying service.

Bandwidth is scalable and dirt cheap. It's last mile that causes cable MSO's their capacity problem and Verizon doesn't have that problem.

Neither cable or telcos have a backbone throughput capacity issue.

LOL all you want about bandwidth being "DIRT cheap". Ciena is the most popular uplink for the PON systems. You are looking at 4 racks of this system to haul the traffic of 40 PON systems over a range of 20 Mi. Dump $700k for that alone, JUST the transport cost that much. That's $550 per subscriber just for the backbone transport 20Mi long, add about $100k per 20Mi extension up to 4 extensions.

Verizon spent thousands per sub for deployment.

You think $550 per sub over the lifetime of that equipment is a lot? You don't understand Verizon's EOS then.

$550 is NOTHING to them. How do you think they get to a number like $21,000,000,000 for deployment by 2010?

Residential HSI is not going to break the back of Verizon's new network. Not by any stretch of the imagination.

Dirt cheap right now is about $10/mbps wholesale. Bottom line: it's all about the ratio of access:head-end bandwidth. Based on their current pricing, Verizon appears to be looking for somewhere in the ballpark of 10:1 to 8:1 for edge:head-end bandwidth ratios. If heavy use applications continue to gain popularity and degrade that ratio significantly you are either going to see price hikes or usage-based billing. You can't sell $150 worth (15mbps) of wholesale bandwidth + all the local connection fees (delivery/support infrastructure) for $65.

They serve dozens of subs on a per Mbps basis. So while it costs them $10, they gets hundreds in revenues.

Even with heavy use applications, you will NEVER have everyone using their connections at the same time.

With P2P applications that generate traffic *constantly*, concurrency is a guarantee. Also, for video applications you're going to have a peak time from 6pm - 10pm, with the 7-9pm block being the heaviest hit.

You think P2P is a problem for Verizon?

It's only a problem for cable because a few seeders are capable of saturating upstream channels. You actually think that MSOs are going after BT because they have backbone connection capacity issues? No way. Otherwise EVERY provider including telco DSL providers would be capping right now.

To compare, DOCSIS 1.1 (the primary standard used by Comcast, among others) provides 38mbps, on widely varying node sizes. So if you have a node of say, 200, which doesn't seem outlandish, that gives you 194kpbs per person... FiOS and DOCSIS aren't even in the same realm with regards to oversubscription.

Not sure how many times I have to say this but YES FIOS IS BETTER.

However, EPS my post was in response to Skeedatl who on multiple different threads claims the following
"400Mbs to each home"
Never been more wrong

"Will never have a capacity problem"
And my Daughter will never date, I WISH

"Docsis upstream serves 1000 customers in some cases"
Cable Co's ALWAYS use multiple upstream ports per downstream port on the CMTS.

"So long as they don't raise speeds over much over 100Mb, their topology isn't suceptible to oversubscription."
My point was to prove this wrong.

I don't have a problem with facts and the fact is FiOS has the advantage when it comes to last mile connection. However, I do have a problem with misinforming people

All of that which I stated is absolutely true and you have yet to show otherwise.

Verizon's network topology on the local level isn't suceptible to overselling and on the macro level is totally scaleable.

The demand you guys are touting that would even start to cause Verizon a problem isn't reality...no where even close.

GPON = 2.488Gbs
Number of subscribers = 32 or 64
Mac bandwidth without over subscription = 77Mbs or 38Mbs

"So long as they don't raise speeds over much over 100Mb, their topology isn't suceptible to oversubscription."
and
"400Mbs to each home"
BOTH WRONG!

A direct quote from on of the CMTS manuals...
"Recommended max CMs per line card = approximately 1000-1200 modems per line card primarily dictated by worst case download speeds customer wishes to tolerate for its subscribers during peak activity time.

Recommended max CMs per US receiver = Not greater than 200 per upstream port primarily dictated by return path Noise,SNR, collision level control."

which means that this statement is...
"Docsis upstream serves 1000 customers in some cases"
WRONG!

Only time will prove this one for me.
"Will never have a capacity problem"

Lies are what I'm arguing here, not the fact that FiOS is better than Docsis. FiOS is better than Docsis but that's ne reason to make false claims about FiOS.

But FiOS is better, FiOS is better don't you know that FLengineer???? FiOS beats Docsis 3.0

Sorry I felt the need to reply to my own post because I know it's comming...

Yeah but after many days of very clear explaintions what you don't seem to grap is why it's better and what it means in terms of Verizon's capacity. It's a quantum leap in capacity compared to cable and DOCSIS 2.0. And what you don't seem to get is that cable is just now running into capacity issues with DOCSIS 2.0 and you think Verizon is on the cusp of having these same capacity issues.

You seem stuck on backbone connectivity which isn't a problem for cable or the telco. The bottle neck for cable MSOs or telcos ain't the backbone connection.

Since I know it's coming...

64 subscribers 64 subscribers 64 subscribers...

No guy, it's per home (whether they subscribe or not) and there is no provider that has a 100% subscription rate nor 100% of their subscribers using 100% of their connection 100% of the time that's why Verizon wouldn't have an overselling problem even though the math made indicate it to be possible. It's also why G-PON allows for 400Mbps residential. P2P piracy isn't that popular (MSOs claim these users are 1-5% of their subscriber base) but with DOCSIS 1.1 and 2.0 it only takes a few P2P pirates to saturate the upstream channels and that is why they make news and give MSOs headaches in some of their franchies. P2P just isn't the bandwidth consuming nightmare you think it is.

DOCSIS 3, SDV and ditching analog solves that problem for cable, so much so that Roberts is bragging about a 100Mbps future.

These DL caps have nothing to do with HSI traffic. It's about killing off PPV VOD competitors before they do to cable what Apple and Wal*Mart did to the music industry.

That's not even the same paradigm with the music industry. This isn't an issue of whole albums & physical media vs digital downloads of single tracks. We're talking about VoD through an Internet company that has to pay for the content, their own head-end servers and bandwidth, to be connected to an ISP that has to pay for Internet bandwidth plus infrastructure to deliver that content to a user. It's that solution in contrast to the provider's own VoD solution which is Content servers -> ISP internal network -> subscriber.

When you factor that the broadband providers are using more efficient video delivery mechanisms on bandwidth that doesn't directly interfere with their broadband service offering, of course they are going to have an advantage.

Regarding your arguments on the capacity front, while Verizon clearly has the better infrastructure right now with FiOS, to say they will NEVER have capacity problems is just short sighted. Everybody likes to keep quoting the poor jerk that said "640KB ought to be enough for everybody". Never is a long damn time.

Of course it's an issue of whole albums and single tracks.

It's the money and competition. VOD competition means cable can no longer charge what the market would bear but must start matching VOD competitors.

No one said that 640KB is enough for everybody for ever, it was enough for the time.

Verizon can't run into a capacity issue unless a lot of things happen. First, Verizon's technology stops being scaleable and given advancements in fiber transport that is unlikely. Second, Verizon must have a competitor emerge that pushes Verizon to capacity. Cable even with DOCSIS 3 isn't capable of pushing Verizon to capacity.

So while theoretically Verizon could be pushed to capacity if some magic elf releases 1Gbps NLOS WiFi for $5/mo but in reality there is nothing that can come even close to competing with Verizon's shiny new $21,000,000,000 network, not now or even on the horizon.

IOW, it's comes down to possible vs probable. Could Verizon eventually reach the capacity of their infrastructure? Possibly? What is the probability given the situation that would have to happen for FTTH to not live up to residential demand? Slim.

Even if that distant future brings pure ultra-HD 2560x1600P IPTV with 50-150Mbps streams...what could the residential demand possibly be? 1Gbps? 1Gpbs decades from now? Between now and then that demand could be met using the most expensive parts of Verizon's current infrastructure. Now don't get me wrong...I'm not saying that every screw and card will last 500 years...I'm talking about the most expensive components. Over time Verizon will incrementally update hardware and both ends of of the fiber just like MSOs do with DOCSIS but only if competition drives them to it.

Sigh...I've already been through this a zillion times with you.

Number of subscribers 32 or 64

WRONG - Number of HOMES 32 or 64. Unless you think Verizon is going to convert 100% of DSL, dialup and steal every single cable sub, that will never be the case.

400Mbps to each home, has been demonstrated by Verizon and is certainly possible given consumption trends of residential subscribers. Again, for the zillionth time...G-PON offers 4X the capacity of DOCSIS 3 split between only a tiny fraction of homes and only now is cable having an issue with BT seeders saturating upstream channels.

CMTS manuals...recommended...

Wrong: You actually think every single MSO in every single instance follows recommendations? Which means yet again you are wrong. You have no clue about cable MSO topology or how they're stretching DOCSIS resources.

Verizon will NEVER have a capacity problem.

Last comment because hopefully I have convinced the readers that FiOS is not everything Skeedatl makes it out to be, here's a QUOTE FROM VERIZON

"But even with network capacity upgrades, oversubscription, and multiplexing, signing up for 100-Mbit/s service won't mean you'll get that bandwidth all the time. "100 Mbit/s isn't guaranteed," O'Byrne says. "If you ran a test on the network, it would read 100 Mbit/s, but if enough people were demanding it, we couldn't guarantee it."

Duh, of course they say that...they say that in their advertising.

To say otherwise would be an SLA and Verizon doesn't offer SLAs on their residential service.

They key there is "if enough people were demanding it"...they'll never have enough people demanding it. They're already charging $140/mo for 30Mb in the majority of their markets. You think there is big demand for it? Go take a look at the DSLR Archive...there isn't.

Yeah, if they get a 100% subscription rate and 100% of those users pony up $140 for their top tier and 100% of those users using 100% of their connection, Verizon could have an issue. But that is not reality.

Here's a few simple questions that if you answer honestly, will show you why Verizon won't have an oversubscription problem even if they offered a 100Mbps service, even 400Mbps service.

1. What do you think is a realistic FiOS HSI subscriber rate for Verizon given cable and DSL competition? Meaning what realistic percentage of those 32-64 homes would actually sign up for FiOS HSI?

2. What percentage of those users buying FiOS HSI would pony up $100-$150 a month for a 100Mbps service? (most FiOS markets charge $140/mo for 30Mb service)

3. What percentage of those users spending that amount of money would run their connections wide open 24/7/365.25?

I think none of these would ever be even close to 100%, and as such Verizon will NEVER have an oversubscription problem. They simply have too much capacity and the services are expensive. There isn't enough cable HSI competition, even with DOCSIS 3, to push Verizon to offer speeds their network couldn't support.

All I can say to you Skeed is to keep reading about the technology and watch what you type because you give false information quite often. I never said they would have an over subscription problem, I simply stated that it is in fact oversubscribing because you said "with out oversubscription". Also, Verizon will have to upgrade their system because they will max out their current capability, It will take awhile for that to happen but it will happen. a333 mentioned WDM-PON and EPON, currently one is to expensive and one has trouble with TDM traffic but they will work those problems out and eventually Verizon will have to upgrade their GPON system to one of these. I realize this doesn't mean that they have to lay more fiber in the ground but it is still an upgrade because of needed bandwidth.

Also, Like most providers Verizon uses the subscription rate to tell them how many HOMES to pass. Verizon DOES pass around 50 homes with a fiber that serves 32 because they want to get as many as they can on each node.

We're talking about 2 different types of oversubscription.

There is mathematical oversubscription, that is the total possible consumption assuming 100% subscription and 100% use of a maximum tier compared to total capacity. This is what you seem to be referring to.

There is actual oversubscription, that total consumption based on actual number of subscribers and the actual tiers they are subscribing to compared to total capacity. This is what I'm referring to.

YES - I agree that in a worse case senario, it is mathematically possible for Verizon to oversubscribe their service if they deploy 100Mbps service which is why Verizon doesn't offer an SLA.

However, I believe that Verizon will never reach a point of actual oversubscription. There are simply too many limiting factors that prevent Verizon from overselling a network with such high capacity which is why I stand firm on my statement that "their network isn't suceptible to oversubscription". These limiting factors are also why I stand firm on my statement that Verizon could easily support 100Mbps and 400Mbps tiers without caps and without oversubscription.

To support 100Mbps dedicated service they only need to sell 5Mbps vs 100Mbps at a ratio of 3:1, even with a 100% subscription rate on a 64 home GPON deployment. Meaning to dedicate 100Mbps for a customer they need only sell 5Mbps to 3 other customers. Verizon is doing WAY better than that ratio because of the high prices they charge for those high end tiers (see chart above as indication of what plans people are buying).

Looking at subscriber rates and competition I just don't see Verizon oversubscribing G-PON even if they offered a 400Mbps tier, again, because few could afford to buy it and those that did can get dedicated bandwith thanks to the sale of so many 5Mbps subs. They would only need to sell 5Mbps to 13 subs for every 400Mbps sub that buys. Again, they do way higher than that 13:1 because of their pricing.

IOW, Verizon's network is not suceptible to oversubscription because of their insanely high capacity, real world subscription rates (the percentage of those 32 or 64 homes that actually buy FiOS HSI), and tier buying trends as a result of Verizon's pricing schedule. The total number of actual subs X those subs' tiers won't exceed Verizons available capacity, thus they aren't oversubscribed.

Also, many here seem to think that GPON is the end of the road for FiOS. We still haven't gotten to WDM-PON and EPON. Those systems can use the EXISTING physical FiOS plant to easily deliver speeds far beyond the realms of cable or DSL. As to backbone, they're already considering 40 Gbps equipment, AND have the capability of going for 100 Gbps cards using the existing fiber, only replacing the line cards on either end. Besides, DWDM has only started off. Right now, I believe it's around 30-50 wavelengths that it handles, but R&D firms all over are churning new tech every day, so 1-Tb links may not be such a daydream. Also, Verizon doesn't have the peering issues suffered by other companies, as they themselves are a Tier-1 carrier with shitloads of capacity.

Ciena CoreStream supports 80 wavelengths in the C-Band. L-Bank is yet to be touched but available.

you can say things like EXISTING all you want but remember that only about 45% of that network actually EXIST right now.

But in no part of the network (except MAYBE some backbone links) is there noticeable congestion issues. DSL? It is shared at the CO-backbone level, plus it is relatively low bandwidth. FiOS? 2.4 Gbps split among 32/64 units. So no last-mile congestion issues here.
As to backbone, as you've said, there is even 80-wavelength DWDM equipment out there that can even further use backbone fiber already in place. So, where's the bottleneck?

What's going to happen is that people will start choosing Internet service based on cap, rather than on instantaneous speed. What's the point of having a 50/20 connection if there is a cap of, say, 100GB? If you're willing to pay the extra money for a fast connection, chances are you're going to hit the cap.

I downgraded my service to the slowest possible speed just for this reason. Instantaneous speeds don't matter all that much to me when I'll just hit the cap if I use them anyway.

The thing is, unlike bottom end telco plans (like Verizon's $15-$20 DSL) the bottom cable plans are typically still pretty pricy thus highly profitable. The cable operators would love you to pay $45 for a 6Mb plan you never use.