Funding hitch to plans for 'next generation' network

By RICHARD WOOD

Plans to build a "next generation internet" network capable of speeds of up to 10 gigabits per second (Gbps) for New Zealand science and academic research are bogged down in funding problems that are unlikely to be resolved until next year's budget, at the earliest.

The initial target of a September launch was not met and the initiative has transferred from those who know the technology to those who can bend a political ear.

Meanwhile scientists fear New Zealand is falling further behind the international research community.

The original NGI consortium changed tack in May to become a registered society able to enter into legal and funding arrangements and take a longer term role.

But the baton of driving the concept has effectively passed to an advisory group of scientists headed by Dr Peter Hunter of the University of Auckland, which is seeking funding from a group of government departments. These include the Ministry of Research Science and Technology, the Ministry of Economic Development, Tertiary Education Commission, NZ Trade and Enterprise and Treasury.

The scientists have completed a report on what is needed and will go back to Science and Technology Minister Pete Hodgson in the next couple of months.

MoRST strategic planning manager Andrew Kibblewhite said the report would show New Zealand was about five years behind where it should be.

Scientists were spending wasted energy getting around the lack of the network and were not able to send and receive data as regularly as their international colleagues.

In the meantime the Government came to the party early this month in a small way by allocating $200,000 for a business case study on the concept. It has been estimated that $10 million is a starting figure for a NGI network.

Chief executive of the NGI-NZ society, Tone Borren, said the need was urgent for a network, which should have an excess of bandwidth able to take the incremental and huge spikes of activity that research projects could generate, without incurring massive charges under traditional telco business models.

Borren said the reason the consortium had aimed at September was because a lot of infrastructure already existed, such as fibre optic cabling.

Borren said that for a while the group thought a big bang approach would be possible but now he believed a pilot was required first.

A key technical change was now being proposed that would force more competitive pressures into the network.

Borren said the "Gigapop" nodes in each centre where local users would link into were now envisaged as neutral operations where previously this was not necessarily going to be the case.

The aim with the pilot was to ideally set up six Gigapops - one in each university city at a cost between $2 million and $3 million.

The core network needs to be able to deliver speeds variable between 1Gbps and 10Gbps and this cost would need to be covered by supply contracts to the user organisations.

The Herald understands this network and the cost of connecting overseas are where the real high costs come in. Researchers are shipping disks around because it is cheaper.