Sergii Figurnyi – Stock.adobe.co
Given Labour’s ambition to use AI as a driver of economic growth, there are questions to be answered about how the nation’s energy grid will cope.
AI Energy Council holds its second meeting in order to discuss how to upgrade the UK’s electricity grid to support the power required to run the next wave artificial intelligence-powered applications.
Most technology providers are aiming to deliver advanced AI capabilities in their product portfolios. This will increase demand for AI in the datacentres, and lead to an increase in energy consumption. Kate Brandt, the chief sustainability officer of Google, released its 10th annual sustainability report. The report shows a 27% rise in electricity demand at Google’s datacentres. Brandt said that despite the rise in energy consumption, Google has been able to decouple its operational energy growth and associated carbon emissions. She said that this was achieved by adding 2,5 GW of clean electricity to the grid.
At the AI Energy Council, which brings together energy companies, technology companies, the energy regulator Ofgem, and the National Energy System Operator, the meeting aims to assess how stakeholders can collaborate to forecast the energy required to deliver a 20-fold increase in computing capacity over the next 5 years.
As part of the Industrial Strategy, the government has committed to supporting its 50-point AI Opportunities Action Plan that was announced in January. The Action Plan acknowledges the need to expand the UK’s computing capacity so that scientists and AI companies in the UK can access the systems they require to develop AI-based system.
It is expected that the meeting will also cover which sectors will adopt AI quickly and how this may lead to significant shifts in energy consumption. Discussions will center on the assumptions that need to be made in order to accurately forecast demand and ensure the energy system is ready for AI.
Peter Kyle, Secretary of State for Science, Innovation and Technology, said that giving our researchers and innovators the processing power they require will not only maintain the third-largest AI power in the world, but will also put British expertise at heart of the AI advances which will improve our life, modernise public services and spark economic growth, which is the cornerstone to our Plan for Change.
We are clear-eyed on the need to ensure that we can power this golden age for British AI using responsible, sustainable energy sources. Today’s discussions will help us to drive forward this mission, delivering AI Infrastructure which will benefit communities across the country for generations without ever compromising our clean energy superpower aspirations.
One of the areas discussed is the UK’s energy requirements to deliver a 20-fold increase in computing capacity over the next 5 years. The meeting will also cover which sectors are most likely to adopt AI quickly and how this can lead to significant shifts in the energy demand, as well as the future energy requirements of the AI sector.
The Department for Science, Technology and Innovation (DSTI) said that once final sign-offs are in place from energy regulator Ofgem, more than 400GW of extra capacity could be released from the grid connection line, which would provide electricity for AI projects the government views as essential to economic growth.
Caroline Donnelly,
By Caroline Donnelly.
Datacentre energy deficit: How worried should operators be about power outages?
By Chris Holmes.
By Caroline Donnelly.
Datacentre energy deficit: How worried should operators be about power outages?
By Chris Holmes.
