{"id":45605,"date":"2026-05-22T12:17:16","date_gmt":"2026-05-22T12:17:16","guid":{"rendered":"https:\/\/www.europesays.com\/britain\/45605\/"},"modified":"2026-05-22T12:17:16","modified_gmt":"2026-05-22T12:17:16","slug":"isambard-ai-building-the-uks-most-powerful-ai-supercomputer","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/britain\/45605\/","title":{"rendered":"Isambard-AI: Building the UK’s most powerful AI supercomputer"},"content":{"rendered":"

\"\"<\/a>image: \u00a9University of Bristol
\n Simon McIntosh-Smith looks at Isambard-AI, which combines highly advanced systems architecture and rapid deployment with world-leading energy efficiency to deliver UK\u2019s most powerful AI supercomputer<\/p>\n

Supercomputers are the foundation of modern research and innovation. They power groundbreaking projects in areas critical to our lives, such as climate modelling, drug discovery, Large Language Model (LLM) training and AI security, enabling researchers to process vast datasets and simulate complex systems at an unprecedented pace.<\/p>\n

Isambard-AI is the UK\u2019s most powerful supercomputer. In November 2025, it ranked 11th most powerful in the world on the TOP500, the international league table for private and public compute.<\/p>\n

Delivered through the Bristol Centre for Supercomputing at the University of Bristol, Isambard-AI forms a central pillar of the UK\u2019s AI infrastructure ecosystem. It is a core component of the AI Research Resource (AIRR), funded by the Department for Science, Innovation and Technology (DSIT) and provides the compute for Sovereign AI, funded by the UK Government.<\/p>\n

Following launch in July 2025, access to Isambard-AI opened to academia and industry via AIRR; with the launch of the Sovereign AI fund, this further extended to start-ups and innovators. This ensures that the supercomputer\u2019s power is distributed across the UK economy.<\/p>\n

A system built for AI<\/p>\n

Isambard-AI has been designed specifically for AI workloads, particularly large-scale model training and data-intensive simulation.<\/p>\n

The system is built on the HPE Cray EX platform, delivered in partnership with Hewlett Packard Enterprise. At its core are 5,448 NVIDIA GH200 Grace Hopper superchips, each tightly coupling a Central Processing Unit and Graphics Processing Unit into a single architecture optimised for AI.<\/p>\n

These are organised into 1,362 compute nodes, with four superchips per node. Each node operates with a unified memory space of approximately 864 GB, which is critical for efficiently handling large AI models without excessive data movement.<\/p>\n

Across the full system, this delivers over 21 exaflops of AI performance, enabling large-scale model training at global-leading scale; around 200 petaflops of traditional high-performance computing performance, supporting scientific workloads; and nearly 25 petabytes of high-performance storage, designed for data-intensive
AI pipelines.<\/p>\n

The system is connected using HPE Slingshot interconnect technology, providing high-bandwidth, low-latency communication across nodes.<\/p>\n

Efficiency as a consideration<\/p>\n

Data centres are well known as being large users of electricity, powering both the hardware and cooling systems. The power used for the 5MW facility is all purely renewable, from UK sources. Isambard-AI uses direct liquid cooling (DLC), removing the need for fan-based systems and significantly reducing energy overhead.<\/p>\n

This sends a mix of coolant and water around a closed loop, which runs in small pipes through the blades. Heat is passed from the blades to the liquid, and this is taken away to be cooled outside before the coolant is delivered back into the system.<\/p>\n

The system is in a modular data centre using HPE\u2019s MODPOD design, which enabled rapid deployment while reducing construction emissions. There\u2019s no building to speak of, no toilets or offices. Isambard-AI sits inside what looks like four large shipping containers, attached together, protected by rigorous security defences.<\/p>\n

This approach means:<\/p>\n

A build approach that cut construction emissions by approximately 72% compared to traditional data centres
\nPower usage effectiveness (PUE) is roughly 1.06, meaning 6% of the total energy used to power the computer is used additionally to power the cooling system. Traditional systems have PUEs of around 1.9 to 2, meaning they use 90% to 100% additional power for the cooling system.<\/p>\n

The project has proven that when building AI infrastructure, performance and sustainability can be engineered together. As a result of all of this design and thought, Isambard-AI launched at #2 on the Green500 list of supercomputer efficiency.<\/p>\n

From funding to full system in under two years<\/p>\n

Traditional supercomputers of this scale typically take three to five years from funding to deployment. In contrast, Isambard-AI moved from concept to operation in under two years, with some phases delivered in under 18 months.<\/p>\n

The timeline is unusually compressed:<\/p>\n

Late 2023 \u2013 Government investment confirmed through DSIT and UKRI under AIRR.
\nEarly 2024 \u2013 System design finalised and site build started.
\nMid 2024 \u2013 Physical installation begins using modular data centre infrastructure.
\nLate 2024 \u2013 Early users access the system for pilot workloads.
\nMid 2025 \u2013 Full system operational.<\/p>\n

This pace reflects a deliberate shift in how national infrastructure is delivered, moving closer to the speed of industry than traditional public-sector programmes.<\/p>\n

AIRR and Sovereign AI<\/p>\n

Through the AIRR, Isambard-AI is being positioned as a national resource for all. Access is allocated through structured calls, prioritising projects that deliver clear scientific, economic, or societal impact.<\/p>\n

AIRR enables broader access across academia, industry, and the public sector; allocation is based on merit, impact, and alignment with national priorities, from healthcare to climate science.<\/p>\n

Through Sovereign AI, funded with \u00a3500 million from the UK Government, portfolio companies are provided with dedicated access to Isambard-AI. This addresses some of the most significant constraints in AI development, the cost of compute and the infrastructure required to innovate.<\/p>\n

Training advanced AI models requires access to supercomputers optimised for AI workloads. Without infrastructure at this scale, many companies simply cannot compete, and many start-ups go abroad to scale.<\/p>\n

By integrating compute access into funding packages, the Sovereign AI programme is lowering the barrier to entry for frontier AI development in the UK, linking public investment directly to commercial innovation.<\/p>\n

Successful applicants gain access not only to the system itself but also to expertise from the Bristol Centre for Supercomputing, helping them optimise workloads and scale effectively.<\/p>\n

Applications at scale<\/p>\n

Isambard-AI is helping British Heart Foundation researchers to model proteins in the quest to identify treatments for cardiomyopathy; training UK LLMs and including minority ethnic UK languages such as Welsh, Gaelic and Cornish; identifying skin tone bias in skin cancer detecting apps; understanding how to best train AI agents to ensure they are secure and how easy they are to corrupt; modelling pollution to within 1sqm of the whole of the UK, and identifying cattle who may be displaying signs of sickness to support agricultural efforts. This is a UK national treasure which will benefit the world.<\/p>\n","protected":false},"excerpt":{"rendered":"image: \u00a9University of Bristol Simon McIntosh-Smith looks at Isambard-AI, which combines highly advanced systems architecture and rapid deployment…\n","protected":false},"author":2,"featured_media":45606,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_share_on_mastodon":"0"},"categories":[4],"tags":[2096,4687,23020,5,6],"class_list":["post-45605","post","type-post","status-publish","format-standard","has-post-thumbnail","category-uk","tag-artificial-intelligence","tag-contributor-profile","tag-oag-051-july-2026","tag-uk","tag-united-kingdom"],"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@UnitedKingdom\/116618189480111054","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/posts\/45605","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/comments?post=45605"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/posts\/45605\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/media\/45606"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/media?parent=45605"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/categories?post=45605"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/britain\/wp-json\/wp\/v2\/tags?post=45605"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}