Will the NZ$900 million Pacific Fibre cable mooted by messieurs Drury, Morgan, Tindall, Rushworth, Humphrey and Wiggs succeed? Certainly, the list of founders is impressive with everyone on it not only having an understanding of why Big Fat Pipes with Low Latency are a good thing, but they also have real incentives for the Pacific Fibre cable to succeed - Rod Drury and Sam Morgan especially, with their Internet-based businesses depending on good networks.
John Humphrey is a satellite and cable guy of long standing that I met at the Shin Sat IPstar launch way too many years ago (and, I ended up with a huge dish on my roof as proof that it was indeed possible to deliver decent broadband via birds high up in the sky).
Mark Rushworth and Lance Wiggs get what the Internet is about, and having Stephen Tindall onboard isn't going to hurt Pacific Fibre in the slightest.
Of course, having lots of well-known names onboard doesn't equate success but Pacific Fibre is launching into a market that's difficult to satiate at the moment. It's almost as if you need to dimension networks according to the capacities of their computers' storage and interface speeds - you need to get as close as possible to those two factors to keep customers happy.
The 5.12 Terabits per second that the Pacific Fibre cable is supposed to launch with probably isn't that much over the next decade or so, and it's good to see a 12 Tbit/s upgrade has been factored in.
Having lots of bandwidth for large video, audio, and picture files is one thing but I found it interesting to talk to Rushworth today, and hearing him emphasise low latency as a competitive point of difference. Mark's right of course. NZ and Australia are a huge distance away from everywhere, so the less packet delay, the better. From memory, the best roundtrip you get over Southern Cross is 120 to 130ms to LA, which translates to 160 to 170ms for end users in Auckland. Not disastrously high, but if you can shave 20 to 30ms off that, people with real-time apps are going to be much happier.
Now, the money: $900 million's not chump change, and finding the funding will be a challenge for Pacific Fibre. Given the stated goals of Pacific Fibre, I think this is one project that the government should consider investing in. Morgan's said Pacific Fibre isn't about maximising profits, but to build infrastructure that will benefit the entire country.
Kordia, a state-owned enterprise is said to be teaming up with Pacific Net, but I think a direct investment in the system by the government may be in order, provided the business plan is sound. For starters, it would give our academic network KAREN another less expensive option to Southern Cross for international transit and maybe even get them close to the twin 10GB wavelengths that the Aussie AARnet has.
Some years ago, I spoke to a MED mandarin who said that the then government was given the option of buying a ten per cent stake of the Southern Cross Cable. If that had happened, things may have been very different now, but the government didn't take up that option. Here's a second chance, and Key's cabinet should take it.
Maybe there's a way to micro-finance the cable too, with individuals communities in NZ and Australia chipping in small amounts but in large numbers?
I'm also wondering if Southern Cross smelt something in the wind last year, when its sales director Ross Pfeffer started talking about the possibility of a new cable or upgrading the existing system. The SCCS has been a great success for Telecom, Singtel/Optus and Verizon and it'll be interesting to see how they respond to the Pacific Fibre initiative.
Price cuts on the SCCS could happen, but they won't be much use unless we see some seriously improved pricing on national delivery as well. My ISP contacts tell me that Auckland-Wellington can cost more than Auckland-LA nowadays. Maybe Pacific Fibre needs to have a think about national delivery as well?
Update some industry and academia comment that arrived via Science Media Centre:
Brett O'Riley, CEO of NZICT, comments:
"As New Zealand only has one high-capacity link into the country today, I know major global data centre operators have not considered NZ as a suitable location, despite the fact that we have lower power prices, and greener power, than most of the other countries bidding. But we haven't had redundant capacity and the low cost structures people are looking for.
We can see that international prices out of New Zealand are more expensive than our trading partners, and we've been advocating for a decrease in those prices and we see competition as being the best lever for that.
"From a science perspective, potential projects like the Square Kilometre Array (SKA) will be generating 'astronomical' data sets - it's estimated the SKA will generate more data in a day than the whole world does in a year, currently.
"So we see this new cable and other planned initiatives as being absolutely essential if New Zealand and Australia are to make a compelling case to host the SKA, and clearly there will be many other spin-offs and opp for the science community from this.
"It will also have an extremely positive impact on the ICT sector, given the improved latency which will enable New Zealand companies to offer software as a service remotely to major economies, and will enable New Zealand to continue its growth as a development shop for other digital economies."
Dr Vladimir Mencl, e-research services and systems consultant at the University of Canterbury comments:
"The existing KAREN network has solved the problem of connections between research organisations. But because of tight data caps, universities use a cost recovery model for internet access in their commercial operations, often with costs charged back to the home department of the university. As a result off the KAREN network, I will often spend a lot of time looking for alternative ways to download a file more efficiently.
"The KAREN network has been over-built intentionally so there is a lot of capacity for video-conferencing and it offers low-latency services. So I don't think a new cable would offer much there, there isn't a great strain on the network. He have international collaborations using the BlueFern supercomputer at Canterbury with other countries including Australia and Canada as well as local research organisatons.
"But KAREN doesn't really have the model right yet. It doesn't run purely on a cost-recovery basis, it is still centrally funded by the Government. If cheaper international bandwidth means it is cheaper to run KAREN that's money that could be saved and diverted to other research projects."
Update II Google might want to put some dosh into the Pacific Fibre cable too. Probably a more likely investor than our straitened government.
Other related posts:
Video: Kim Dotcom and Mathias Ortman at the IITP Mega breakfast
Two-factor authentication broken
The problem with naming and shaming
comments powered by Disqus