Gekco
Expert Member
Good question.
Apart from that what about all the areas where ADSL is unavailable because the network has been built on Fibre?
Whether be via a DSLAM box or Fiber to the premises (FTTP)...
-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: Will you be subscribing to the new Disney+ R49 streaming package?
Backhaul bandwidth costs
- Thread starter rpm
- Start date
Good question.
Apart from that what about all the areas where ADSL is unavailable because the network has been built on Fibre?
Whether be via a DSLAM box or Fiber to the premises (FTTP)...
Gekco
Expert Member
So, in conclusion... The existence of a broadband price break is an illusion. These articles are spaced in time skillfully to sow confusion and help those who are guilty of Wholesale broadband price collusion...
Right?
Or something like that...
Does seem that way...
Drunkard #1
Expert Member
- Joined
- Aug 15, 2007
- Messages
- 3,668
- Reaction score
- 18
Those prices are reasonable, so I'll cross JINX off my list.
The last big Q (for "local only" access) is, "What do the big ISPs charge to peer?"
If a Wisp (no IP Connect charge) has enough customers, then local backhaul is insignificantly cheap per Gig (less than 1c). Last mile charges are low (maintenance only), so the only significant charge is the peering charge. What do the likes of IS and SAIX charge to get that Gig off their servers. Not the news servers, but the servers that their clients are paying them for space on, and paying them for bandwidth to. Do the big ISPs expect to be paid twice for the bandwidth? Otherwise local bandwidth should be virtually free! (and international should be only slightly more costly.)
OK, so ADSL users pay for IP Connect, but that's the only charge keeping prices high. We should focus our protests there. Can't we lay an official complaint with ICASA or the Competition Commission?
AndrewAlston
Group Head of IP Strategy – Liquid Telecommunicati
I read these articles with great interest, because I find they are often very isolated segments of the whole picture. And the huge mistake that is often made is not to look at the entire picture at once.
First of all, the cost of putting fiber in the ground has very little to do with the cost of the cable itself, 99% of the cost is in the ducting and actually trenching that ducting into the ground.
Secondly, the fiber game is kind of unique, if you are leasing fiber to a company, and your entire business is fiber, you have to cost to be sustainable. Now keep in mind that the chances are, that customer is not going to come back to you for "upgrades", so your costing model has to cater for this.
It also needs to be said that metro based fiber is more expensive than long haul fiber, its FAR more expensive to lay fiber in the metro area than it is to pull in long haul, primarily because long haul trenching allows for far more use of mechanized trenching, in the metro when you're digging up pavements and roads, its often done by hand.
I can say beyond a shadow of a doubt, having fiber in the ground for high speed connectivity if you own or lease that fiber cuts costs, by a *HUGE* amount. Let me give you guys an example, as everyone knows, we had DFA pull fiber from Mtunzini landing station into Durban. Total fiber distance was approximately 152 kilometers. Now, there are a number of pairs there, but we are using a single pair at the moment (no need to use the others at the moment). Total bandwidth on that pair at the moment... *30 GIGABIT* (3 10gig wavelengths, used for various things). The cost of upgrading that bandwidth... R300k per 10G until we hit around the 200gigabit mark, after that it would cost me a million or two to put more kit it to take more line cards. We would max out at 1.6 TERABIT on a pair of fiber. Work out the costs per megabit at those speeds, and its a joke.
Yes, fiber is going to cost the smaller guys a lot of money to lease it/own it, and that is one of the reasons why I like the idea of a second company that owns fiber and does low cost leasing of circuits (155mbit circuits or something to that effect), because its no secret that for a small player, owning the fiber isnt necessarily economical. But for the big guys, who are the guys actually charging the customers, and even for large corporates, the fiber model that DFA has makes so much sense.
I have to make another comment about DFA while I'm at it, their service is nothing short of exemplary, and considering the turn around time to fix fiber faults that they have, its worth every penny. I know of two instances where DFA has had to deal with breaks, and has had them fixed in under 4 hours in both cases. That is nothing short of astounding when you consider what is actually involved in fixing broken fiber, and I might add, their 4 hour SLA they offer on metro fiber is *FAR* superior to what you will find if you buy fiber from most of the international companies. Also, before people whine about costs, look at the cost of pulling fiber in London, in New York, or in most international locations. Trust me on this, the prices we're seeing are pretty good by comparison
Just my 2c
Andrew
First of all, the cost of putting fiber in the ground has very little to do with the cost of the cable itself, 99% of the cost is in the ducting and actually trenching that ducting into the ground.
Secondly, the fiber game is kind of unique, if you are leasing fiber to a company, and your entire business is fiber, you have to cost to be sustainable. Now keep in mind that the chances are, that customer is not going to come back to you for "upgrades", so your costing model has to cater for this.
It also needs to be said that metro based fiber is more expensive than long haul fiber, its FAR more expensive to lay fiber in the metro area than it is to pull in long haul, primarily because long haul trenching allows for far more use of mechanized trenching, in the metro when you're digging up pavements and roads, its often done by hand.
I can say beyond a shadow of a doubt, having fiber in the ground for high speed connectivity if you own or lease that fiber cuts costs, by a *HUGE* amount. Let me give you guys an example, as everyone knows, we had DFA pull fiber from Mtunzini landing station into Durban. Total fiber distance was approximately 152 kilometers. Now, there are a number of pairs there, but we are using a single pair at the moment (no need to use the others at the moment). Total bandwidth on that pair at the moment... *30 GIGABIT* (3 10gig wavelengths, used for various things). The cost of upgrading that bandwidth... R300k per 10G until we hit around the 200gigabit mark, after that it would cost me a million or two to put more kit it to take more line cards. We would max out at 1.6 TERABIT on a pair of fiber. Work out the costs per megabit at those speeds, and its a joke.
Yes, fiber is going to cost the smaller guys a lot of money to lease it/own it, and that is one of the reasons why I like the idea of a second company that owns fiber and does low cost leasing of circuits (155mbit circuits or something to that effect), because its no secret that for a small player, owning the fiber isnt necessarily economical. But for the big guys, who are the guys actually charging the customers, and even for large corporates, the fiber model that DFA has makes so much sense.
I have to make another comment about DFA while I'm at it, their service is nothing short of exemplary, and considering the turn around time to fix fiber faults that they have, its worth every penny. I know of two instances where DFA has had to deal with breaks, and has had them fixed in under 4 hours in both cases. That is nothing short of astounding when you consider what is actually involved in fixing broken fiber, and I might add, their 4 hour SLA they offer on metro fiber is *FAR* superior to what you will find if you buy fiber from most of the international companies. Also, before people whine about costs, look at the cost of pulling fiber in London, in New York, or in most international locations. Trust me on this, the prices we're seeing are pretty good by comparison
Just my 2c
Andrew