The United Kingdom has long been a leader in technical innovation and today is home to a wealth of innovative tech companies from an emerging set of startups to established companies that have been a part of our lives for decades. These companies are moving faster, pushing further, and uncovering more as the power and potential of high performance and intense computing applications open doors—or at least open them more quickly. These companies are tackling everything from the natural language processing found in smart home IoT devices, to the artificial intelligence (AI) now used in connected, autonomous cars.
Advanced analytics and high-performance computing (HPC) applications are creating opportunities for enterprises every day. Nevertheless, these new applications, and the volumes of data they create, are causing companies—and the executives within them—to reconsider how data is stored, accessed, and most importantly, powered. Perhaps for the first time in recent history, the ability to realize the full potential of future technologies might be gated by the access and availability of our previous generation’s power grids.
The report, Tech Nation 2017, which analyses the digital tech ecosystem in the United Kingdom, found that the digital technology sector grew 50% faster than the overall economy. Yet, as the country prepares to play a leading role on the world stage in delivering real business value from these transformative technologies, the environmental impact of this form of energy consumption must also be evaluated.
The abundance of power needed to support the data produced and analysed by ‘computing’ broadly is staggering. For example, the amount of electricity used by data centers globally in 2016 was roughly 416 terawatts, which is greater than the entire UK energy consumption in the same year, and this number is only expected to increase. Similarly, the volume of data that just HPC applications alone produce is estimated to grow by 40 zettabytes by 2020. For comparison, the entire amount of information stored online in the year of 2013 consisted of only four zettabytes of data. An exponential increase added by the data generated by HPC applications will add to the power challenges that Europe as a whole already faces.
Electricity consumption rose across Europe in 2017, making it the third consecutive year to increase. Given that 70% of continental Europe relies on the same grid, this increase of new power consumption will be felt everywhere, as cross-border power outages are increasing at an alarming pace. In addition, spare energy capacity in the UK now stands between 6.2% and 8.2%, and is likely to shrink even more. With an ageing power grid that already is working near capacity, any increase in power consumption will amplify the number of outages, which have been on the rise since 2013. The UK also relies on predominantly fossil fuels to produce energy. Therefore, as the adaptation of intensive compute applications increases—in addition to the data centers needed to power and house them—this will in turn lead to the production of more greenhouse gases. For computer-aided manufacturers, researchers, and financial institutions reliant on HPC to drive innovation at this furious pace, HPC clusters must find a home that provides sustainable, renewable and cost-effective energy. How can this be accomplished?
Currently, the financial services sector in London is faced with market conditions and regulatory pressures that have pushed firms to focus on analysing data faster to arrive at smarter decisions in order to deliver services more quickly. As a result, firms rely on HPC to improve the speed and accuracy of their results. HPC has become a vital part of the financial services landscape, as various AI techniques are used to identify trading signals, model strategies, conduct real-time risk analysis, and compute complex pricing calculations. Challenged to keep pace with transactions, innovations and progress within the hyper-competitive financial markets, the IT divisions of banking and asset management companies are being pushed to get the most out of their power investments. For this reason, the need for hyper efficient and super reliable HPC, not only at scale, but at the lowest cost as possible, is integral.
It has become standard practice for many UK-based organisations, like financial institutions, to store their HPC applications in high cost data centers in the middle of London or other urban business districts. For latency insensitive compute, this practice is illogical and inefficient as not only are the data centers powered on some of the most expensive and unsustainable energy around, but many of the HPC applications that are being stored in these data centers are non-latency specific—eliminating the need for their storage to be in close proximity to the staff utilising the applications.
Consequently, many forward-thinking innovators in the UK are looking to Nordic countries like Iceland to manage their compute-intensive applications due to its advantageous and clean power profile. Having one of the world’s most reliable power grids, Iceland offers long-term, low cost, massively abundant and scalable energy, which together provides impressive savings across the total cost of operation. No wonder that a number of UK-based asset management firms are already using the benefits of Iceland for their big data analytics and computational research. Iceland’s power grid is rated among the most reliable in the world, powered by dual sources of renewable energy: hydroelectric and geothermal.
As enterprises in the UK start to rely on computationally intensive workloads and HPC applications more and more, previous generation’s power grid infrastructure in the UK could restrict or slow down future technological progress. For this reason, companies must be forward-thinking and utilise other, more energy rich and sustainable locations for their computing needs, and the answer may lie in Iceland.
Adam Nethersole, Senior Director of Marketing at Verne Global
Image Credit: Lassedesignen / Shutterstock