Rethinking Icelandic Data Center Opportunities

Iceland has been chasing the data center industry for almost four years. Although the promises of the large scale data centers have not come to fruition, there is hope for this industry. There are still, however, some significant challenges that must be overcome.

At the outset, the benefit to locating data centers in Iceland was the use of 100% green power, the cost of that power, and the ability to use outside air for the majority of cooling. As we look at it today, technology progress within the industry, and aggressive programs in other countries have lessened some of these advantages. The combination of improved air-based cooling systems and increased data center operating temperatures now make free cooling is practical in many countries. To remain competitive, Iceland needs to emphasize and validate the value of green power and make sure that its pricing is as good or better than competing countries.

The biggest challenge at the outset was the cost and reliability of the network. That is still a challenge. My view was that this should be taken out of the equation and that access to Icelandic data centers should be provided through hand-offs in London, Amsterdam, and other locations. Done properly this would make Icelandic data center offerings more competitive, less challenging, and, hopefully, lessen concerns about reliability.

Over the last four years, other challenges have emerged including natural risks from volcanoes and earthquakes, the financial crisis in Iceland, and unfavorable tax status for certain users of data centers. The natural risks are mitigated through proper design and siting and the other risks can be solved through legislative efforts and careful structuring of contracts for future users and developers.

The most positive developments in Iceland are the emergence of new Icelandic data center and cloud providers. These are small scale companies, but they are actively dealing with many of the issues above. I believe that there are opportunities for more companies like this and that they can create more significant employment opportunities than the mega data centers. I also believe that the mega data centers will come, but that they are several years and many challenges away.

Will Current Volcanic Activity Impact Iceland's Viability for Data Center Services?

Last week Rich Miller wrote a piece in Data Center Knowledge pondering the impact of the recent volcanic activity on the potential for data centers in Iceland. Although certainly these events create the potential for negative perception, it is important to realize that (i) they are not historically unusual, and (ii) have been accounted for in every aspect of siting and design of both the data centers and supporting infrastructure. Here are some additional comments that I posted in response to the article:

The current volcanic activity is not so much an anomaly. Eyjafjallajökull, the volcano that is currently erupting is along a belt of high geophysical activity. Several years ago we heard of a new island rising from the ocean off of Iceland. This is the same belt of activity. It is distant from population centers and viable data center locations. This region and it phenomenon are taken into consideration in the design of all electrical and communication systems. In fact, the newer cables land west of this region so that they will not be separated from the population and data centers should there be an event such as this. The older cables have alternate routes that traverse the northern edge of the country.

The issue of potential (data center equipment) contamination in Iceland or other countries needs to be monitored, but should not be much of a concern. Prudent designers using air side economization typically use a heat exchanger or high efficiency filters to trap vehicle emissions and other fine particulate. Although volcanic fallout may foul this equipment and increase maintenance costs it is highly unlikely to result in any material damage to equipment within data centers.

It's All About the Network

Yes, Iceland is an island in the middle of the North Atlantic. Through history this has proven to be an immense measure of protection and a throttle on outside influences. One wonders if and how the pervasive connectivity brought in the modern day will change the country.

The modern age of communication in Iceland started in 1994 with the construction of the CANTAT 3 cable. This cable had a capacity of 7.5 Gbps and allowed Iceland a more stable and larger connection than had ever been achieved through previous cable, satellite, and radio links.

CANTAT 3 was supplemented by FARICE 1 in 2004 with 100x the capacity, although only a small portion of that capacity has been put into service.

In recent years, Iceland has pushed to expand network capacity even further with the installation of "Danice" and "Greenland Connect" as well as study of additional cables to the US and Ireland.

Unfortunately, these cables are very expensive to construct. Since the capital costs are large, the cost of capacity is also very high. So there is something of a conundrum-the capacity exists to connect Iceland to the world and support growth of new industry, but the economics still need some work.

In the post 2001 telecom bust, many cable operators went bankrupt leading to a re-marking of assets. This, combined with a glut of capacity drove the cost of network between New York and London to a level that promoted vast changes in how data is stored and transmitted. In the end, this drove tremendous network growth and finally brought the cable operators to profitability.

This is what needs to happen in Iceland. They need to figure out how to deliver capacity to Europe for less than 2,500 EUR and to the US for less than $5k per 10 Gbps wave. If these points can be met, they will open up a broad range of business possibilities involving Iceland.

So the secret is to develop pricing models based on ultimate capacity and utilization and not on early cost recovery.