UK triples funds for AI supercomputer to £300 million to boost AI safety research


UK triples funds for AI supercomputer to £300 million to boost AI safety research



The UK government has announced a significant boost to its artificial intelligence (AI) prowess, with a tripling of its previously announced £100 million investment in its ‘AI Research Resource.’

The £300 million investment, as the Department for Science, Innovation, and Technology reported, is set to bolster the UK’s most advanced computer, Isambard-AI, along with a newly announced Cambridge supercomputer, ‘Dawn.’ The investment follows the country’s public commitment to AI safety as it hosts the world’s first-ever summit on AI safety at Bletchley Park.

Aiming to enhance AI capabilities and foster breakthroughs in various fields, including drug discovery and clean energy, the investment will significantly increase the capacity of the UK’s current largest public AI computing tools.

Isambard-AI, stationed at the University of Bristol and built by Hewlett Packard Enterprise (HPE), will be equipped with 5,000 advanced AI chips from Nvidia, making it ten times faster than the UK’s current fastest machine.

To grasp the processing power of the supercomputer, consider this: it will deliver over 200 ‘petaflops,’ executing 200 quadrillion calculations per second. This dwarfs even the fastest smartphones today, managing only trillions of calculations per second.

Perhaps even more staggering – a human would need to make one decision every single second for 6.3 billion years to match what this supercomputer can calculate in one mere second. Its processing capabilities eclipse anything else available in the UK currently, with speeds 10 times faster than the nation’s swiftest existing machine.

Simultaneously, the Cambridge-based ‘Dawn’ supercomputer, a result of a partnership between Dell and UK SME StackHPC, will be powered by over 1,000 water-cooled Intel chips to minimize power consumption. Scheduled to be operational within the next two months, ‘Dawn’ will target breakthroughs in fusion energy, healthcare, and climate modeling.

The UK’s Frontier AI Taskforce, chaired by Ian Hogarth, will be given priority access to these state-of-the-art computing tools. The focus of the task force’s activities will be to analyze the safety of advanced AI models to mitigate the risks posed by AI, including concerns around national security due to the potential development of bioweapons and cyberattacks.

This initiative was announced by the Science, Innovation, and Technology Secretary Michelle Donelan at the AI Safety Summit at Bletchley Park. She emphasized that

“Frontier AI models are becoming exponentially more powerful. We have made it clear that Britain is grasping the opportunity to lead the world in adopting this technology safely so we can put it to work and lead healthier, easier and longer lives.”

This strategic move, she continued, will ensure that the UK’s leading researchers and scientific talent have the tools they need to explore and understand this complex technology.

The first-ever AI safety summit has seen ministers from around the globe emphasizing the need to ensure the safe and ethical development of AI. The summit highlights the international concern and urgency to mitigate the potential risks of AI that surpass human abilities, underscoring the need for such substantial investments in AI safety.

The UK’s investment into these supercomputers provides a tangible demonstration of its commitment to lead in the area of AI safety as the world grapples with the implications of AI’s rapid advancement.

As these supercomputers come online in the coming years, the forefront of AI safety research and analysis is set to see a significant acceleration.



Source link