How the UK can sustainably develop high-performance computing and artificial intelligence

In May 2024, the UK joined the European Joint Undertaking for High Performance Computing (EuroHPC), which brings together supercomputing resources from 35 countries, including Norway, Turkey and all 27 European Union (EU) member states.

By joining the prestigious program, the UK aims to strengthen its leadership in science and technology, support international+ activities and use HPC to drive innovation and economic growth – all of which represent hugely positive ambitions for the UK IT industry. However, membership of EuroHPC will also accelerate demand for energy in the UK, and this may even exceed the country’s current capacity. Especially considering that the UK has yet to cope with the huge increase in energy demand resulting from the increased use of artificial intelligence (AI).

UK National Grid chief executive John Pettigrew has commented that artificial intelligence will consume 500% more energy in the UK over the next decade. This is not just a UK problem, but this spike in voltage is quickly becoming a global problem. Data centers around the world consumed 460 terawatt hours (TWh) in 2022, representing almost two percentage points of total global electricity demand. Due to increased energy demand from artificial intelligence, electricity consumption in EU data centers is expected to be 30 percent higher in 2026 than in 2023.

According to a recent study by Goldman Sachs, artificial intelligence will increase data center power requirements by 160% by the end of the decade. This isn’t unexpected, considering Google’s emissions have almost doubled in the last five years thanks to AI. These shocking numbers raise questions about the feasibility of achieving net zero emissions by the end of the decade.

The key question remains how can we sustainably meet this level of demand?

Youlian Tzanev

Co-founder and Chief Strategy Officer of NexGen Cloud.

The biggest trends of our time

Artificial intelligence and climate change are two of the most important and influential stories of our time, each significantly shaping different aspects of society, the economy and the environment. They are far from mutually exclusive. Artificial intelligence will undoubtedly play a role in tackling the challenges of climate change. However, data center energy demands will put pressure on achieving net zero targets.

Like the need for electricity, the UK also needs access to GPUs as it looks to become a leader in HPC and AI. However, innovations must be sustainable to ensure energy-intensive processes can be carried out with minimal impact on the environment and without compromising the UK’s energy security.

100% sustainable data centers

The only real solution is to use renewable energy to power data centers.

Energy-saving strategies in AI data centers also need to be prioritized to enable the UK to keep pace without having to pull the plug due to energy demand. Acquiring renewable energy to power data centers is associated with several challenges, primarily related to the need for constant and reliable power supply, high initial investment costs and the integration of renewable sources with existing infrastructure.

Some renewable energy sources, such as solar and wind, are intermittent in nature, requiring the development of efficient energy storage solutions and grid improvements to ensure a constant supply of energy. NexGen Cloud operates an AI Supercloud in a data center on the outskirts of Oslo, which uses hydropower to power and cool its infrastructure. Hydropower is continuous, allowing the data center to be powered by clean energy around the clock.

Investment and sovereign clouds

Market forces alone will not be enough to get us there. The UK government’s direct investment in GPU chips is also key to growing the country’s HPC and AI industry.

Government investment in GPU infrastructure is essential because current funding lags significantly behind other global players. UK investment in Nvidia GPU chips dwarfs orders from tech giants such as X and Elon Musk’s China. According to Pitchbook, $122 billion has been invested in generative AI companies in the US, while European generative AI companies have received a relatively modest $3.8 billion. Such disparities make it difficult for the UK to compete in the HPC and AI sectors, requiring a strong government push to increase GPU availability and infrastructure investment to bring it up to international standards.

Additionally, adopting sovereign cloud solutions can help UK businesses gain access to high-power GPUs while ensuring compliance with data protection regulations, thereby increasing the UK’s competitiveness in the global AI market.

Cloud infrastructure is already in place to ensure that AI operations can remain compliant with European data sovereignty and privacy laws, keeping them within European jurisdiction. By powering our data centers with 100% renewable energy, the UK and EU can significantly mitigate the environmental impacts associated with high performance computing and artificial intelligence, thus enabling innovation to continue without jeopardizing the ability to achieve net zero by 2030.

Essentially, this could provide both a blueprint for how the UK can sustainably support HPC and AI and become a true world leader in these sectors.

Introducing the best eco-friendly hosting.

This article was produced by TechRadarPro’s Expert Insights channel, where we highlight the best and brightest minds in today’s technology industry. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in cooperation, you can find more information here: https://www.techradar.com/news/submit-your-story-to-techradar-pro