Australia's NBN is in the process of being rolled out, and I was recently asked to comment on whether or not Australia had missed a golden opportunity by Phil Morle at Pollenizer. The post below is a slightly edited and expanded version of my response.
First, I think it's worth looking at what this means, not just for Australia's technoratti, but for the wider population and what it could do for Australia's economy. Much ink and many photons have been spilled discussing whether or not the NBN is even a worthwhile endeavour. Why $43B? Where's the business case? What's the ROI?
Why so Much Early Scrutiny?
First an opinion, I think it's worth looking at why and how the NBN, or more specifically NBN Co, will be heavily scrutinized early on. Thanks to the big dollars involved, Mike Quigley and his team are in a race to show progress and will be judged quickly on how they execute on the following
1) How many homes are passed?
2) How much did it cost? (And how much was Telstra paid)
However, what we're interested in, is "What will people use it for?"
NBN Co must deliver on 1) and 2) before the demand issue will come to the fore.
The NBN is about the last mile and providing choice. The NBN provides services up to layer 2 with Retail Service Providers (RSPs) providing layer 3 services and above. It's from this design choice that the opportunities will flow. With RSPs afforded the choice of what layer 3 services to provide they are free to offer IPv4, IPv6, multicast (the NBN will provide some help here at layer 2 to make this possible), anycast, and jumbogram services.
This means that we will see the emergence of applications that will exploit the capabilities of these services.
The IPv6 Opportunity
IPv6 provides a whole heap of improvements over IPv4 that include larger address space, zero-configuration, and security. RSPs that offer IPv6 will not need NATs; this means that we can have HEAPS of wireless devices that can communicate directly with each other AND move between different parts of the network. Start thinking what it means when tablets like Apple iPad and Microsoft's Surface are everywhere. There's also no reason why everyone's home based WLAN AP couldn't be made available for public use. I think we'll see federations of WLAN APs involving multiple providers emerge.
The Multicast Opportunity
Multicast has been the next big thing for many years, I did my Ph.D. in it back in the late 90's. NBN's PON architecture will be great for distributing highly popular content. Some care will need to be taken to ensure that people can source multicast efficiently as well as consume it. Expect to see the mass migration of anything broadcast on to the NBN in short order. Anyone holding a broadcast TV or radio license needs to understand this now since the value of these licenses will effectively be zero in less than 10 years time.
The Datacenter Opportunity
Jumbograms are designed for shipping really large amounts of data around. Normal ethernet packets are 1500 bytes in length jumbograms can be up to 64k long (I think). Data centres are getting bigger, SAAS is taking off, Ultra HDTV, 3D TV, will be delivered via the net, all of these things rely on the ability to ship huge amounts of data between replication points. The existence of a population with access to real broadband and backbone that can efficiently really huge amounts of data around will make Australia an attractive place for people to not just try out new applications but to host them as well.
I think we'll see some very interesting things happen in Australia with massive data centres opening up near sources of cheap power (Hydro Power Stations in NSW, or Tasmania anyone? What about solar powered facilities up north?) that will serve not only Australian populations but Asia as well. Australia's first-rate privacy laws will also make it attractive place to host from. To encourage the construction of these data centers and a workforce that knows how to use them, I'm think there's a great opportunity here for state governments.
Australian State-Governments have a history of providing incentives to targeted industries to establish and maintain facilities and operations in key areas. At the very least I think it would make sense for them to provide incentives for the big players in cloud-computing (Salesforce, Microsoft, Google) to establish sizeable data centres here as well as the staff to build services on them. Growth could further be stimulated by doing this under the proviso that excess capacity is made available at peppercorn rates for bona-fide research organizations (Universities, NICTA, CSIRO) as well as for SMEs developing new cloud based services. This approach would ensure that both the infrastructure and skills are available to make a difference.
Finally the establishment of a thriving data centre industry in Australia would have an interesting side effect. Australians currently suffer high bandwidth prices since we download more data into Australia than we pump out. The establishment of a large data centre industry here has the potential swing the pendulum the other way, to significant reduce the inbound / outbound ratio, and perhaps even drive it to the point where we originate more traffic than we consume internationally. Were this to occur the peering charges levied against Australian ISPs would fall dramatically and drive further consumption, and provide ample opportunities to develop new offerings based around the wilful waste of bandwidth (see Abundance vs. Scarcity).
The Workforce Opportunity
Longer term, Australia has a great opportunity to outsource high-value services to the rest of the world. We're a services-based economy in a stable part of the world and enjoy a great lifestyle. I think we'll see people move here for the quality of life and work globally with minimal travel. While some work has been done in this area inside of CSIRO and the Smart Services CRC, much, much more remains to be done. The combination of the NBN and Australia's relatively unscathed emergence from the GFC will make Australia the place to do this work.
So back to the original question.... quick, real initiatives? Real yes, quick no.... I think we'll need at least 18months to 2 years post the inevitable Telstra deal and a critical mass of homes passed before we'll start to see substantive initiatives emerge. Changing the world always takes longer than people expect. That being said the time to start thinking and experimenting is now.
And one last closing shot.... some people question whether or not what's been proposed will be future proof and we'll need to upgrade the NBN. The short answer is only a little bit. Pulling the fiber is the expensive part of the roll-out. Swapping units out of rack is relatively cheap. The architecture proposed will be able very upgradeable and allow for Gigabits/sec to be pumped in each home in the long run. Once rolled out Australia will have a national infrastructure asset that will last for at least a couple of decades.