With an ever-growing population, tech companies are trying to find efficient ways to keep up with increasing data usage, writes Paul Budde.
I HAVE BEEN INVOLVED in the data centre market since the 1990s. However, at that stage we didn’t use the term “data centre”, we talked about "value-added networks".
This development described at an early stage the integration of telecommunications and IT technologies. However, it was not until a full decade later that this market started to take off at a time when several technologies started to mature.
The commercialisation of the internet was the big driver in the second stage of what was by then developed into data centres. The ecosystem that evolved concentrated on three core businesses: telecoms, data management and cloud computing. While all three markets are individually still rapidly evolving, we also see a merger between the three elements.
While the telecoms players stood still during the 1990s and early 2000s, the digital giants (AWS, Microsoft, Google and Apple) continued to develop their infrastructure. As they were largely frustrated by the lack of affordable co-operation from telecoms companies, they basically started to build the infrastructure on their own. They now all operate their own large-scale fibre optic networks and have developed very sophisticated "cloudware" (software). These operations are now known as "hyperscale" data centres.
They offer robust, scalable applications and storage facilities to both individuals and businesses. This is based on massive hyperscale computing and these centres are the size of several football fields. They hold thousands of servers each linked to ultra-high-speed fibre optic networks. With their continued growth and high levels of sophistication, they will keep on challenging the existing I.T. and telecoms companies.
These hyperscale centres are essential in our ever-increasing complex society and economy. What has rapidly become the major problem area for these centres is the extremely high use of energy.
They are desperately looking for solutions to address the ever-increasing costs of energy and the increasing taxes/penalties and other government fees and levies linked to reducing their carbon footprints. Building undersea centres and even outer space centres are all options under investigation.
These developments are driven by the massive growth in data and this is set to continue based on an ever-increasing level of sophistication needed for our societies and economies and both telecoms networks and I.T. developments are assisting with that. Complexity, capacity, reliability, redundancy, scale and costs are favouring the development of these hyper centres.
As these centres will be able to offer services at very competitive prices they will increasingly become a threat to the traditional data centres. At the same time, however, there are other forces that drive the growth of smaller centres such as regional developments, nationalism and cybersecurity.
These hyper centres are a global phenomenon and are reaching well beyond national borders, national markets and national industries, with all of the jurisdictional, taxation and control complexities attached to this.
Their physical locations will depend on low energy opportunities, high level of technological advances, political and economic stability. Until a few years ago, these developments were truly global. Unfortunately, the current political situation will see that these centres will be based in the U.S. for U.S. companies, in China for Chinese companies and most likely also a separate development for Europe.
Countries such as Australia, with the potential of very low energy costs and a more politically neutral position, could potentially also play a global leadership role. That is if the Government is able to develop a future vision for Australia in areas such as energy, telecommunications, innovation and so on.
Hyperscale centres are already a combination of big data, cloud and telecoms and a range of developments in these areas such as edge computing, low latency technologies, IoT (5G) and mesh computing are all important developments that will further assist in the future development of this web of industries.
But it is not all plain sailing. Governments are becoming increasingly nervous about the influence of these digital giants and in coming years we will see an increase in legislation and regulation. Add to this the distinct political trend towards nationalism (instead of globalism), this will put pressure on the giants. They potentially will either be curtailed, broken up or otherwise limited in their expansions.
This could slow the trend towards hyper centres. We also will see more direct government intervention in the development of data centres as well as government investments to safeguard whatever they think is needed to be safeguarded in the name of national interest. How this will pan out depends on the individual ideologies of the countries involved and this uncertainty makes investments in hyper centres at the moment risky; it will all depend on the business model they use for their plans.
In the end, the technical developments we are seeing are unstoppable as well as the world growing towards nine billion inhabitants. The current levels of complexity are only a fraction of what to expect in the coming decades. We will need hyperscale developments to do all of this and, whether we like it or not, we will have to be global in our approach. It will be very interesting to see how we as humanity are going to manage this.
Paul Budde is managing director of Paul Budde Consulting, an independent telecommunications research and consultancy organisation. You can follow Paul on Twitter @PaulBudde.
Support independent journalism Subscribe to IA.