Of course.
-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: Google Maps vs Waze vs Apple Maps vs physical paper maps?
SEACOM reveals bandwidth usage on cable
- Thread starter rpm
- Start date
Of course.
eltherza
Expert Member
Another way to look at it:
1Mbps line can download about 10GB in a 24hour day (someone want to double check my maths please?)
But, lets be realistic, lets say for only half the day you are going at full speed. So, in a day at 1Mbps you can get about 5GB. So for a month, thats about 150GB a month
For seacom, thats about R54 a month. Or about 36c a GB
If we apply our model lets x4 for backhaul, and x4 for ISP it comes to about... R6 a GB
1Mbps line can download about 10GB in a 24hour day (someone want to double check my maths please?)
But, lets be realistic, lets say for only half the day you are going at full speed. So, in a day at 1Mbps you can get about 5GB. So for a month, thats about 150GB a month
For seacom, thats about R54 a month. Or about 36c a GB
If we apply our model lets x4 for backhaul, and x4 for ISP it comes to about... R6 a GB
This is why I still say SEACOM should start their own ISP in SA. the cable is there, so why not use it.
Charge a low price for Uncapped, and stimulate the market - make people used to fast uncapped, make it the norm, and watch as the other ISPs buy bandwidth like crazy to compete.
Seacom - if you are not going to do something like this, our ISP's will keep raping us, and they won't need to buy bandwidth from you.
YOU, SEACOM, need to get data flolwing through your pipes. that's what you do. if there is no data flowing, you are gonna have problems.
So Start an ISP or make a deal with one ISP (like, charge them a lot less for bandwitdh, than what you usually charge) and STIMULATE the Market yourself.
Yes, i know it sounds weird to charge 1 ISP almost nothing for Bandwidth - but it will bring you 20x the amount of ISP's back to you, that NEED to buy from you to compete in the market.
Charge a low price for Uncapped, and stimulate the market - make people used to fast uncapped, make it the norm, and watch as the other ISPs buy bandwidth like crazy to compete.
Seacom - if you are not going to do something like this, our ISP's will keep raping us, and they won't need to buy bandwidth from you.
YOU, SEACOM, need to get data flolwing through your pipes. that's what you do. if there is no data flowing, you are gonna have problems.
So Start an ISP or make a deal with one ISP (like, charge them a lot less for bandwitdh, than what you usually charge) and STIMULATE the Market yourself.
Yes, i know it sounds weird to charge 1 ISP almost nothing for Bandwidth - but it will bring you 20x the amount of ISP's back to you, that NEED to buy from you to compete in the market.
CambodiaDave
Senior Member
Hi Guys,
You are 100% correct on the issues. That's why we set up DigiChilli.
Feel free to mail me and we'll tell you how to get cheaper international bandwidth using Seacom right now.
You are 100% correct on the issues. That's why we set up DigiChilli.
Feel free to mail me and we'll tell you how to get cheaper international bandwidth using Seacom right now.
MadMailMan
Expert Member
There is no demand for more than 100mb on SEACOM at CURRENT PRICE LEVELS. Even if SEACOM bandwidth was free just the DF costs getting it to CT or JHB would be more than SAIX bandwidth costs.
Remeind me what the ISPs pay for IPConnect again?
we don't need LLU we need the "Telkom" infrastructure, including the local loop, taken away from Telskum and given back to the people who have already paid for it (many times over). Let it be run in the same way as ICANN or some other similar org.
What we need is guavament to own up to the fact that it's their fault that SA is so far behind with DSL penetration and they are perpetuating the digital devide. But with the Interwebz comes knowledge and we all know what knowledge is. Besides the guavament REALLY dosent want the masses to be educated and/or knowledgeable, would be counter productive in their eyes.
/rant off, enjoy the rest of your day.
Remeind me what the ISPs pay for IPConnect again?
we don't need LLU we need the "Telkom" infrastructure, including the local loop, taken away from Telskum and given back to the people who have already paid for it (many times over). Let it be run in the same way as ICANN or some other similar org. What we need is guavament to own up to the fact that it's their fault that SA is so far behind with DSL penetration and they are perpetuating the digital devide. But with the Interwebz comes knowledge and we all know what knowledge is. Besides the guavament REALLY dosent want the masses to be educated and/or knowledgeable, would be counter productive in their eyes.
/rant off, enjoy the rest of your day.
InTheCube
Expert Member
Dude, are you guys up and running as yet? Last time I heard there were problems with IP address allocations, or something else with the network. Have these been resolved as yet? I'm dying to sign up.....
jetlee
Senior Member
I dont get the issues here .. surely 10% of the cable and 60% of the cable is practically the same cost (from a seacom point of view) .. so why not light up 70% of the fibre .. for the same price you charge for 10% with short term contracts .. that will shake up the market .. and allow most ISP's to deliver a uncapped solution for cheap .. then slowly increase the prices as the capacity gets taken up ...
Somehow it seemed simpler in my head
Somehow it seemed simpler in my head
MadMailMan
Expert Member
I dont get the issues here .. surely 10% of the cable and 60% of the cable is practically the same cost (from a seacom point of view) .. so why not light up 70% of the fibre .. for the same price you charge for 10% with short term contracts .. that will shake up the market .. and allow most ISP's to deliver a uncapped solution for cheap .. then slowly increase the prices as the capacity gets taken up ...
Somehow it seemed simpler in my head
Because then they will exhaust their capacity before they can make their money back.
It's like having a 10 lane highway that cost R1000 to build, and charging each car R1 to use it but they only open 1 lane. You want them to open 7 lanes and only charge say 10c. But at 10c those 7 lanes will fill up really fast and become congested long before they would have made back there R1000.
Or do you want them to open 7 lanes at R1 a car but they only have 3 cars using all 7 lanes? Remembering all 3 cars could run perfectly fine on just 1 lane.
I like explaning things using cars but sometimes it just confuses things more. Hope this makes sense tho.
Last edited:
CambodiaDave
Senior Member
Telkom status on Realm says "awaiting activation". I'm tempted to post the address of the account service manager here. It might speed things up. But then again, it might bork everything to a looong death too.
Patience is a virtue - especially on other people's slow bandwidth products.
Tjoker
Expert Member
cant wait...........the "other" people are having a nightmare from TomasW & Kie
ingeon
Senior Member
- Joined
- Apr 11, 2008
- Messages
- 752
- Reaction score
- 4
Nice. Make it international news how pathetic SA telecoms fatcasts are.
It`s all just one big monopoly. I blame the competition commision
It`s all just one big monopoly. I blame the competition commision
ingeon
Senior Member
- Joined
- Apr 11, 2008
- Messages
- 752
- Reaction score
- 4
)eugenevdm
Well-Known Member
Dear Suveer Ramdhani,
You are quite correct. Your service is having a slow uptake because the national backbone networks in South Africa are still pretty much owned by Telkom. Neotel via DFA is laying fibre like crazy but this is a slow process and it might take years for them to match what Telkom has already. In the few select places like city centres where Neotel has already laid fibre they *have* to recoup their investment by asking exuberant prices.
Do yourself a favour phone Neotel and Telkom today and get a quote from them for data over fibre. You will find a striking similarity between their prices. Then approach any Telkom data fibre contract client and you will notice that every time Neotel offers them prices Telkom quickly reduces the price. PS. This is written into their contracts. While you're at it phone DFA and ask them how much it's to lay 5 kms of fibre. It does not come cheap I tell you.
So really what you have in a catch-22. Not enough people is using your service because Telkom is still in the way. The only way Seacom will be able to get more clients is if we have a "third" major competitor that is lying fibre aggressively and who is willing to compete on price. Not Vodacom or MTN please because they are already too rich already and they don't need the money so they don't like competing on price. Perhaps also a strong competitor for DFA because really just having one company in SA with the know-how to lay fibre professionally is a bit dof. As for WACS etc - yes you certainly have a problem because even though you're "open-access" it looks like Telkom isn't so interested in using you. I wonder why?
So in summary:
* You are dependant on National backhaul networks
* National backhaul networks are owned by Telkom
* National backhaul competition is being slowly built by Neotel
* Neotel can't match Telkom prices because 'recoup investment'
* DFA can't charge less because 'recoup investment'
* Neotel doesn't want to start a corproate 'price war' because of investment and the ricochet effect this is going to have on Telkom contracts.
Catch-22 I tell you. What you need is someone like Neotel to take a bigger risk in the corporate / wholesale space in terms of price reduction.
kind regards,
You are quite correct. Your service is having a slow uptake because the national backbone networks in South Africa are still pretty much owned by Telkom. Neotel via DFA is laying fibre like crazy but this is a slow process and it might take years for them to match what Telkom has already. In the few select places like city centres where Neotel has already laid fibre they *have* to recoup their investment by asking exuberant prices.
Do yourself a favour phone Neotel and Telkom today and get a quote from them for data over fibre. You will find a striking similarity between their prices. Then approach any Telkom data fibre contract client and you will notice that every time Neotel offers them prices Telkom quickly reduces the price. PS. This is written into their contracts. While you're at it phone DFA and ask them how much it's to lay 5 kms of fibre. It does not come cheap I tell you.
So really what you have in a catch-22. Not enough people is using your service because Telkom is still in the way. The only way Seacom will be able to get more clients is if we have a "third" major competitor that is lying fibre aggressively and who is willing to compete on price. Not Vodacom or MTN please because they are already too rich already and they don't need the money so they don't like competing on price. Perhaps also a strong competitor for DFA because really just having one company in SA with the know-how to lay fibre professionally is a bit dof. As for WACS etc - yes you certainly have a problem because even though you're "open-access" it looks like Telkom isn't so interested in using you. I wonder why?
So in summary:
* You are dependant on National backhaul networks
* National backhaul networks are owned by Telkom
* National backhaul competition is being slowly built by Neotel
* Neotel can't match Telkom prices because 'recoup investment'
* DFA can't charge less because 'recoup investment'
* Neotel doesn't want to start a corproate 'price war' because of investment and the ricochet effect this is going to have on Telkom contracts.
Catch-22 I tell you. What you need is someone like Neotel to take a bigger risk in the corporate / wholesale space in terms of price reduction.
kind regards,
DFA is by no means the only company who knows how to lay fibre in SA. In fact as far as I know they don't do any of the actual work but contract it out to other companies.
All in all, I don't think there is any shortage of companies who have the skills to lay fibre, just a shortage of companies who have the money to pay them to do so.
nictron
Senior Member
When you put it like this you realize how stupid the whole sector really is!
I have a 512Kbps Diginet line here at work and we are paying? ........wait.......wait......R13 000 per month.
How is that even F(&&$(@ possible!
CambodiaDave
Senior Member
We're almost there. Telkom have handed over their leg. Final configs and we're off! Send "Test Account" to dave @ vukisa dot com for expedited access.
HOT!
Incumbant Greed
Unfortunately incumbants like MTN and Vodacom are more interested in maximising profits for shareholders through offering the bare minimum in network capacity and quality while extorting huge amounts of money from the public who simply lie down, roll over and allow themselves to get robbed. The result is poor quality HSDPA that is not very stable to say the least and the backhaul is rediculous. What good is HSDPA speeds of 7.8MB/sec if they can only exist between you and the cell phone tower?
If Seacom wants to sell their bandwidth they really needed to partner with MTN or Vodacom to do so. Neotel is still a struggling little provider with a limited network and are still experiencing serious problems with their network.
Unfortunately incumbants like MTN and Vodacom are more interested in maximising profits for shareholders through offering the bare minimum in network capacity and quality while extorting huge amounts of money from the public who simply lie down, roll over and allow themselves to get robbed. The result is poor quality HSDPA that is not very stable to say the least and the backhaul is rediculous. What good is HSDPA speeds of 7.8MB/sec if they can only exist between you and the cell phone tower?
If Seacom wants to sell their bandwidth they really needed to partner with MTN or Vodacom to do so. Neotel is still a struggling little provider with a limited network and are still experiencing serious problems with their network.
AndrewAlston
Group Head of IP Strategy – Liquid Telecommunicati
Been reading these posts with some interest, and figured I'd add a coupla points.
Firstly, on the calculations of cost and the recoup of investment, keep in mind that no cable system ever bargains on selling the whole cable right off the bat, so they up the price per segment to the ISPs in order to make sure that even if they sell fractional capacity they can still make their money back (which is a perfectly fair and legitimate business practice). As such, the published price of a 10G on seacom is nowhere near as low as you guys are making out (real cost using commercial list price calculated over 20 years is around $13.5k US per day excluding operations and maintenance, which works out at around R315 a month per megabit) . However, keep in mind that no ISP anywhere sells 1:1 ratios to the home consumer market, and if you calculate even at a ratio of 10:1 you end up with a pretty decent price, and a 1:20 oversell on DSL bandwidth would be almost invisible to the end user. However, you then have to multiply that price by 4 because lets be realistic, at the rate of telecomms advance you want any long term circuit to have paid for itself in 5 years maximum. Still, you're looking at around R130 / mbit input cost at a 1:20 oversell on 10gig (obviously these numbers are totally hypothetical and based on a lot of mumbo jumbo that would need to be more carefully calculated).
Now, lets add in the other costs, DSL IPC bandwidth, even if you are buying LOTS of it and manage some how to get it down to say R1200 / mbit. Lets work on an oversell here of 20:1 as well, thats another R60 a megabit in IPC cost. So we're up to R190 odd on uncapped 1mbit input cost. Now double that to cover running expenses, staff and equipment (and thats conservative) for a raw input cost of R380 per 1mbit uncapped dsl per month.
Now, keep in mind that these figures only work if you have a 1:20 complete oversell of a complete 10gig circuit, that means to make that number work, you need *200 THOUSAND* dsl customers, and it gets a LOT more expensive if you are buying in lower quantities. It's simply not as cheap as it looks on the surface.
(Someone can check my maths, and like I say, these figures are totally hypothetical and based on a bunch of assumptions which are probably far from realistic, so please don't think that I claim this is an accurate summary, it is just meant to give a more holistic viewpoint on what it costs to provide these services)
At the same time, as I've said on here many times, Seacom is an amazing project, they have done wonders for certain segments of the market and the pricing impacts are already starting to be seen (look at the drops in price on dsl packages lately), but it also takes time.
Just keep in mind when looking at figures, that you have to take it all into account, taking one element and not working out the rest provides a very inaccurate picture.
NOTE: Everything said above is said in my personal capacity, and contains my own numbers, theories and assumptions and is in not representative of anything in my official capacity as TENET CTO
Firstly, on the calculations of cost and the recoup of investment, keep in mind that no cable system ever bargains on selling the whole cable right off the bat, so they up the price per segment to the ISPs in order to make sure that even if they sell fractional capacity they can still make their money back (which is a perfectly fair and legitimate business practice). As such, the published price of a 10G on seacom is nowhere near as low as you guys are making out (real cost using commercial list price calculated over 20 years is around $13.5k US per day excluding operations and maintenance, which works out at around R315 a month per megabit) . However, keep in mind that no ISP anywhere sells 1:1 ratios to the home consumer market, and if you calculate even at a ratio of 10:1 you end up with a pretty decent price, and a 1:20 oversell on DSL bandwidth would be almost invisible to the end user. However, you then have to multiply that price by 4 because lets be realistic, at the rate of telecomms advance you want any long term circuit to have paid for itself in 5 years maximum. Still, you're looking at around R130 / mbit input cost at a 1:20 oversell on 10gig (obviously these numbers are totally hypothetical and based on a lot of mumbo jumbo that would need to be more carefully calculated).
Now, lets add in the other costs, DSL IPC bandwidth, even if you are buying LOTS of it and manage some how to get it down to say R1200 / mbit. Lets work on an oversell here of 20:1 as well, thats another R60 a megabit in IPC cost. So we're up to R190 odd on uncapped 1mbit input cost. Now double that to cover running expenses, staff and equipment (and thats conservative) for a raw input cost of R380 per 1mbit uncapped dsl per month.
Now, keep in mind that these figures only work if you have a 1:20 complete oversell of a complete 10gig circuit, that means to make that number work, you need *200 THOUSAND* dsl customers, and it gets a LOT more expensive if you are buying in lower quantities. It's simply not as cheap as it looks on the surface.
(Someone can check my maths, and like I say, these figures are totally hypothetical and based on a bunch of assumptions which are probably far from realistic, so please don't think that I claim this is an accurate summary, it is just meant to give a more holistic viewpoint on what it costs to provide these services)
At the same time, as I've said on here many times, Seacom is an amazing project, they have done wonders for certain segments of the market and the pricing impacts are already starting to be seen (look at the drops in price on dsl packages lately), but it also takes time.
Just keep in mind when looking at figures, that you have to take it all into account, taking one element and not working out the rest provides a very inaccurate picture.
NOTE: Everything said above is said in my personal capacity, and contains my own numbers, theories and assumptions and is in not representative of anything in my official capacity as TENET CTO