how-to block ads
|
|
Share Topic
 |
 |
|
|
 espaethDigital PlumberPremium,MVM
join:2001-04-21
Minneapolis, MN
kudos:2
 ·Clear Wireless
| reply to pspcrazy
Re: I think we are seeing how TWC will try to "educate" people said by pspcrazy:Why should they pay more when they're already paying 60 or so dollars a month. That's a crazy amount considering most server companies provide a dedicated server, with a 100 mbit or higher line give you 2 tb for 130 a month. That with the server rental and service they provide. If you lived in a carrier neutral facility taking an Ethernet handoff from your upstream provider in the same facility, then maybe that pricing model would apply. I have a number of friends/colleagues that operate hosting companies in the Chicago and DFW markets, and I can tell you that they technically lose money when people use their full 2TB allocation on a dedicated box. It's a numbers game -- for every person that comes close to hitting their 2TB allocation there are 10 customers who never exceed 100GB on the month.
For a parallel example, think casinos. They can afford to pay out big on occasion because overall they still make money on the average.
said by pspcrazy:If TWC lowered base cost to 10 dollars per month for the line and a reasonable 10 cents per gb for bandwith then we'd be fine with it, but they try to rip everyone off. Everyone had a right to hate caps, it's because the companies behind adding them are doing so with excessive charges. It's really not as excessive as you make it out to be. Yes, there is profit integrated into the pricing; this isn't UNICEF, these are profit-seeking entities. That said, the base cost is going to work out to be more than $10. Think of shared core IP network costs, network engineer salaries (+ benefits), 24x7 NOC, 24x7 HSI-specific call centers/agents, field technicians/vans(maint+gas+depreciation)/tools/training. I wouldn't be surprised if the base price with no profit built in isn't closer to $25-$30.
The bandwidth pricing is also more expensive at the edge because the delivery technology is vastly more expensive.
On a Cisco Nexus 7000 I can get a 32 port x 10GigE line card for about $70k MSRP, now that card is 4:1 oversubscribed so for my argument I'll fairly limit it to 8 full-rate ports. On a uBR10012 (DOCSIS) platform you can get a 24 port DOCSIS3 (bonding line card) for $89,900 MSRP.
$70,000 / (8 * 10,000mbps) = $0.88/mbps for the MSRP hardware cost
$89,900 / (24 * 38mbps DOCSIS channels) = $98.57/mbps for the MSRP hardware cost.
In the data center example above if you need more bandwidth in your carrier neutral facility you give $200-$500 to the facility owner to run some interpositioning fiber for the cross-connect and you can upgrade capacity by just plugging it in.
In the DOCSIS/DSL world (assuming a remote terminal) you need to trench more fiber to the new location of your node / DSLAM and re-engineer the cabling plant around the newly placed infrastructure. Equipment costs aside, people doing this work don't do the job for free. They demand competitive salaries so they can pay bills, feed their families, and have a quality of life just like everyone else.
said by pspcrazy:Btw electricity is finite don't compare it to bandwith which is essentially unlimited. Electricity isn't finite -- you can have as much as you want, as long as you're willing to put up the cash to get it.
said by pspcrazy:IF RR was really smart they'd even allow local lan peers (same line) to transfer at line speeds but they aren't because it doesn't make them any more money then they currently are making. There's no such thing as a "LAN peer" on a DOCSIS plant. Everything goes through the CMTS by design, whether you are sending to your neighbor's house or out to the rest of the Internet. The last mile commonality is the same in either case. | |
| Thank you very much for at least putting some numbers in your statement. | |
iansltx
join:2007-02-19
Golden, CO
kudos:2 Reviews:
 ·Comcast
| reply to espaeth
Thanks for the hard numbers. However seems as though the cablecos are using technology that isn't quite suited for the purpose if they're paying that much per megabit. Heck, you can get a 34-ish Mbps WiMAX base station for around $9k these days. More expensive than cable? Yes, but it's wireless 
Additionally, for your pricing you're looking at a one-time cost for the equipment. Amortize it over twelve months and the numbers look much rosier. Or over 36...probably closer to the life cycle of the equipment. Correct me if I'm wron on this.
What's even funnier is how Comcast doesn't seem to have a problem upgrading their systems to handle more bandwidth and customers, their caps are halfway decent (I'd prefer 500GB over 250GB myself but ah well) and their business tiers are within reach of the average Joe if they want to pull more than a few hundred GB in a month. In contrast, where's TWC's $90 16/2 biz tier? | |
espaethDigital PlumberPremium,MVM
join:2001-04-21
Minneapolis, MN
kudos:2 Reviews:
·Clear Wireless
| said by iansltx:Thanks for the hard numbers. However seems as though the cablecos are using technology that isn't quite suited for the purpose if they're paying that much per megabit. Heck, you can get a 34-ish Mbps WiMAX base station for around $9k these days. More expensive than cable? Yes, but it's wireless I think if you compare coverage area, the costs work out a little more favorable for staying wired.
said by iansltx:Additionally, for your pricing you're looking at a one-time cost for the equipment. Amortize it over twelve months and the numbers look much rosier. Or over 36...probably closer to the life cycle of the equipment. Correct me if I'm wron on this. You're right on this. The point I'm driving at is people like Karl Bode  like to continually suggest that broadband companies are marking their service up 1000-1500% by referencing a "best-case" scenario price for data center bandwidth. The problem with that calculation is that it completely ignores the very different CapEx and OpEx costs on the broadband side, which is what I was really trying to show. On the datacenter side you can spend $70k (list) and get 80,000mbps, whereas your incremental add on the DOCSIS side is $89,900 (list) to get 912mbps. You can start to see why expansion at the edge becomes a stickier situation.
As for OpEx, I know that you're very familiar with the hosting industry where Karl & co base their pricing. Major hosting providers can get away with having a handful of dual-purpose techs on staff (server/network), and maybe an actual network engineer or two. Contrast that to a broadband provider that needs to operate their own training and certification program (including the cost of trainers & material prep), provide data-centric call center agents, have a fleet of vehicles and field techs for customer calls, field techs for plant management, and head-end engineers / network engineers. To suggest the service and support of a network that spans miles with hundreds of points of distribution / amplification is the same as a colo facility contained to a single server room is a gross misrepresentation of infrastructure costs.
Yes, there is some inflation of pricing to bolster profits -- I don't think anyone is disputing that. At $1/GB overages it's not razor thin margins, but at the same time it's not the highway robbery that some on this forum seem to believe it is.
said by iansltx:What's even funnier is how Comcast doesn't seem to have a problem upgrading their systems to handle more bandwidth and customers, their caps are halfway decent (I'd prefer 500GB over 250GB myself but ah well) and their business tiers are within reach of the average Joe if they want to pull more than a few hundred GB in a month. In contrast, where's TWC's $90 16/2 biz tier? Bigger companies are better able to take advantage of the over/under split a little better, and Comcast has already allocated the capital for the DOCSIS3 upgrades. I think right now it's the bigger players who aren't facing a near term (6-18 month) capacity crunch sitting back and letting the other guys take the hit first.
I think TWC's initial numbers were a little too close to the current usage margins, and user reactions certainly show that. I also think that while a capacity crunch isn't hitting today, the landscape has the potential to change pretty rapidly in the next few years. The time to start working out the logistics of these types of billing and growth issues is now -- not 3-4 years from now when your back is against the wall. (that is, of course, unless these companies are trying to set themselves up for some kind of government bailout package in 3-4 years by arguing they couldn't predict a shift in customer behavior) | |
|
