Cable Industry Finally Admits Caps Not About Congestion → Re: Yeah, let's just ignore the access charges

You Obviously know nothing about FIOS and that for all practical purposes Verizon IS a major backbone of the US internet.

Verizon could care less what its users send/receive as being a tier one provider, it costs the same for one bit or one trillion GB, they have no peering charges

Peering charges have nothing to do with the stuff I was thinking about.

Unless Verizon can break the laws of physics, their costs would start increasing exponentially once they start having to stitch multiple 2M$ routers together to accommodate peak demand across network nodes. Having to do this at few strategic facilities is one thing but having to do it systematically network-wide would kick costs up a few notches.

The biggest single-chassis routers can handle about 5Tbps of non-blocking traffic. Once you need to go beyond that while maintaining close to non-blocking routing, things get a whole lot more expensive and that is definitely where Verizon or any other ISP with millions of subscribers would end up if everyone was trying to use 100+Mbps during peak hours, ignoring potential congestion at the DSLAM or node/CMTS level.

People keep saying that equipment gets cheaper but what they almost always neglect to mention is that density in routed Tbps/rack only doubles every ~5 years, which is much too slow to keep up with peak demand which increases by 50-60%/year which is about 10X over the same period. Since technological progress alone is insufficient to meet demand (about 5X too slow), how many identical switches or routers do you think you need to put together to double the capacity of a single one of the same while maintaining non-blocking performance? You need six of 'em... 2X the capacity = 6X the rack space using same-model equipment. The cost scaling is really horrible.

Not every ISP can afford (or is willing) to use the latest and biggest gear available so don't be surprised if there are more stories about smaller ISPs hitting their equipment's practical brick walls in the future or attempting to extract money from their transit providers to cover some upgrade costs.

Then those ISPs can expect to go out of business. and extract money from their transit providers? Please stop making me laugh.

Most incumbent ISPs have pretty close to an effective monopoly over their respective service areas so going out of business due to not upgrading is unlikely.

As for ISPs wanting to extract revenue from their transit providers, Comcast has tried it with Peer1 and Free is trying it on Google so there certainly are some who are tempted to test those waters. While the scheme may be upsetting for CDNs and transit providers, it actually has a handful of advantages if the ISPs' savings from it are passed down to their end-users, one of them being that it takes most capacity-related costs out of end-users' monthly fees... pay directly for physical access, pay indirectly for your actual usage.

I'm not saying it would on its own but its another nail in their coffin. ISPs in this situation are typically very small and are already struggling as it is.

Actually it is France Telecom/Orange and Google. The amount of money these very large ISPs are extracting from these companies is a drop in the bucket in the bigger picture and I'm very skeptical the ISPs will pass on any savings to the users. These companies never do anything that truly benefits the end user. If they were using the money to directly go towards doing network upgrades then I wouldn't be against this concept so much but that isn't the case. They just want another revenue stream and drag out doing network upgrades as much as possible.