Do We Need Commitment in the Next Gen World?
Had an interesting debate today with a long standing ally in the Next Gen Access movement.The discussion centred on whether CIR (Committed Information Rate) is relevant in a Next Gen World, or whether instead the capacity of eNdGAme FttH infrastructure is sufficient to allow effective QoS (Quality of Service) delivery for time-sensitive applications, eg. voice, video conferencing.
An interesting conclusion is the following truism:
“Given a fat enough pipe, all traffic is bursty”
On reflection CIR is evidence of the obsolete scarcity-from-abundance paradigm so beloved of the 20th Century Telco
- in the sense that the need for a commitment to deliver service is proportional to the risk of unavailability of said service…
If service quality interruption is derisked then no commitment is required.
Manage this post
Bookmark this post
Change feature settings
Choose a feature level and image for this post
Post information
Posted 7 Apr 2009
Last edited 20 Apr 2009
Latest revision: 2
Sharing
- Guy Jarvis can edit
- Anyone can comment
Categories:
Communities:
- No communities
Tags:
1. At 8 Apr 2009 10:57, mike kiely wrote:
In todays £20 a month broadband service, the up to 24 Mbps, translates to a 30kbps peak hour backhaul allowance, with a good privider.
FTTC (VDSL or Docsis) moves us to a £40-£50 a month with I assume a good increase in the peak hour backhaul capacity per user.
What are the equivalent for a FTTH world. I am assuming £80-£100 pm - but their is a significant commitment to broadcast TV on this medium in order to make the medium economic.
I am sure how your defining CIR, but different applications need different quality budgets to deliver the user experience expected. A fat pipe at the edge does little to support end to end serice delivery.
2. At 8 Apr 2009 12:57, Guy Jarvis wrote:
@Mike:
In today's leading UK Data Centres, eg. www.manchesterdc.com, the cost per Mbps per month is sub £10
In the FttH eNdGAme world, effectively DC access capabilities are dispersed to the edge of the network, into the First Mile.
The implications of this shift from scarcity to abundance are profound and change the economics of IP supply completely.
Have a look at http://bits.blogs.nytimes.com/2009/04/03/the-cost-to-offer-the-worlds-fastest-broadband-20-per-home/
Particularly this extract:
"Pricing at Liberty varies widely by market. In Japan, its 160 Mbps service costs 6,000 yen ($60) per month. That’s only $5 a month more than the price of its basic 30 Mbps service. In the Netherlands, meanwhile, it charges 80 euros ($107) for 120 Mbps service and 60 euros ($81) for 60 Mbps."
Here in the UK we are not even in the top 100 for FttH and when we start to move from pilot projects to a nationwide push for FttH we will undoubtedly see the same economics apply....
If we are smart about the ownership and governance models for the monopoly infrastructure needed to deliver of course :)
3. At 15 Apr 2009 16:15, mike kiely wrote:
Thanks - good article, so much of the price premium is in 'cost' of losing existing revenue streams.
So an encumbent would follow the £40-£50pm for FTTC and a £100pm for FTTH + installation costs.
I need to check on the US and Japan numbers to see which include and exclude the TV costs.
The challenge for NGA is the word 'affordable'. You need a real reason to go from circa £20 to Cira £40pm and the FTTH will have to include your TV subs.