The Bristol Centre for Supercomputing (BriCS) is a leading research centre based at the University of Bristol, specializing in high-performance computing (HPC) and artificial intelligence (AI). Established in 2024, BriCS hosts some of the UK’s most advanced supercomputers, including Isambard-AI, the UK’s fastest and most powerful AI research supercomputer. Isambard-AI is part of the UK Government’s AI Research Resource (AIRR) program, aimed at boosting the UK’s AI capabilities. It ranks as the 11th most powerful supercomputer globally, 6th in Europe, and 4th greenest worldwide due to its energy-efficient design using zero carbon electricity and advanced liquid cooling technology.
BriCS also hosts Isambard 3, a world-class supercomputer designed for traditional HPC workloads such as clean energy and health research. The centre supports cutting-edge AI research and large-scale scientific simulations, enabling breakthroughs in areas like automated drug discovery and climate research. BriCS is part of a national effort to increase compute capacity in the UK significantly and is engaged in industry and academic collaborations globally. Access to BriCS supercomputing resources is available through competitive application processes for UK-based researchers and institutions.
The centre’s work not only advances scientific discovery but also emphasizes sustainability, with Isambard-AI ranked among the greenest supercomputers globally, featuring innovations like direct liquid cooling that reduces power consumption by up to 90%. It aims to position the UK as a global leader in AI and HPC research, following the legacy of engineer Isambard Kingdom Brunel, after whom the supercomputers are named.linkedin+4
Isambard-AI
Isambard-AI is the UK’s fastest and most powerful AI supercomputer, hosted by the Bristol Centre for Supercomputing (BriCS) at the University of Bristol. It was developed with a £225 million UK government investment and built in close partnership with Hewlett Packard Enterprise (HPE) and NVIDIA.
The supercomputer’s core hardware is based on 5,448 NVIDIA GH200 Grace Hopper Superchips, each combining a 72-core NVIDIA Grace CPU and an H100 GPU. These chips are configured into 1,362 compute nodes with a combined 864 GB of memory per node and ultra-fast interconnects totaling 800 Gbps within nodes and 800 Gbps across nodes with HPE Slingshot 11 technology. Storage includes 20 petabytes of Cray ClusterStor and 3.5 petabytes of VAST ultra-fast solid-state storage.
Isambard-AI delivers over 200 petaflops of performance in traditional HPC workloads and can exceed 21 exaflops for AI workloads operating at 8-bit precision, making it capable of AI research tasks that would otherwise take decades of computation by the world’s population. It is ranked 11th most powerful supercomputer globally, 6th in Europe, and 4th greenest worldwide due to its energy-efficient direct liquid cooling system managing a 5 MW power facility footprint.
The supercomputer enables acceleration of AI breakthroughs in multiple research areas, including automated drug discovery, climate research, personalized medicine, and materials science. It is a vital part of the UK Government’s AI Research Resource (AIRR) program aiming to boost responsible AI development and research capabilities across the UK and internationally.
Isambard-AI was deployed rapidly within 18 months and was phased to go operational starting with an AI testbed environment in early 2024, reaching full capacity by late 2024. It combines enormous computational power and energy efficiency to position the UK as a global leader in AI research infrastructure.bristol+6
Funding
BriCS, hosted by the University of Bristol, is primarily funded and supported through a combination of government grants, national research programs, and strategic partnerships with industry leaders. Its key funding likely includes grants from UK government bodies such as UK Research and Innovation (UKRI) under programs like the AI Research Resource (AIRR) initiative. BriCS also partners with leading technology companies like Hewlett Packard Enterprise (HPE) and NVIDIA, which contributed hardware and technological expertise for the Isambard-AI supercomputer.
Research partnerships for BriCS include collaborations with other UK academic institutions, national HPC centers, and global research entities focused on artificial intelligence, high-performance computing, and sustainable technology development. BriCS is also integrated into broader networks such as the GW4 Alliance of southwestern UK universities and contributes to initiatives that promote AI safety, clean energy, and health research.
Given its strategic role, BriCS collaborates closely with both academic and industrial partners to leverage supercomputing resources for advancing science, innovation, and education. The center’s funders and partners collectively support its goal of positioning the UK as a leader in AI and HPC.bristol+2
Head of BriCS
Simon McIntosh-Smith is a Professor of High Performance Computing at the University of Bristol and the founder and Director of the Bristol Centre for Supercomputing (BriCS), which operates the UK’s Isambard-AI supercomputer. He has a strong background in microprocessor architecture and HPC, having worked in industry for 15 years designing parallel processors and GPUs before transitioning to academia.
He co-founded ClearSpeed Technology, where he was Director of Architecture, helping to develop some of the first modern many-core HPC accelerators. At Bristol, his research focuses on advanced computer architectures, performance portability techniques, and fault tolerance for HPC applications. He is also involved in major national and European HPC projects and has been a key figure in establishing and running the Isambard supercomputing facilities.
Simon McIntosh-Smith is highly regarded as an expert in GPU computing, many-core processors, and sustainable computer architectures. He leads efforts to optimize and deploy large-scale HPC and AI systems like Isambard-AI and is active in advancing HPC research and technology innovation at the University of Bristol.uob-hpc.github+4
The Bristol Centre for Supercomputing (BriCS) employs between 11 and 50 people as of 2025. This includes research staff, technical specialists, and administrative personnel who manage and operate the centre’s advanced supercomputing facilities such as Isambard-AI and Isambard 3. The team is composed of highly skilled experts in AI, high-performance computing, and related scientific research fields, reflecting the scale of investment and technological sophistication at BriCS.
BriCS and ACRC
The Bristol Centre for Supercomputing (BriCS) is closely related to the Advanced Computing Research Centre (ACRC) at the University of Bristol in the context of high performance computing (HPC).
ACRC is the University’s established HPC service provider, delivering CPU and GPU core hours and managing HPC environments for research and teaching. It operates HPC systems like BlueCrystal and BluePebble, used broadly across academic disciplines at Bristol.
BriCS, founded more recently (in 2024), builds on Bristol’s HPC expertise and hosts cutting-edge supercomputers including Isambard-AI and Isambard 3, significantly expanding Bristol’s HPC capacity and enabling pioneering AI research. BriCS benefits from the HPC infrastructure and expertise cultivated by ACRC, integrating these new AI-optimized supercomputers into the broader HPC ecosystem at the university.
Researchers at Bristol have exclusive access to a portion of BriCS resources, and access can also be granted via UKRI national HPC access schemes. Thus, BriCS represents the next generation of HPC at Bristol, running alongside and evolving from the traditional HPC services managed by ACRC, jointly fostering advanced computing for scientific discovery and innovation.bristol+3
Ex-Director of ACRC
Simon Burbidge is a prominent expert in high performance computing (HPC) in the UK. He has a distinguished career encompassing leadership roles in HPC both in academia and industry. Simon was previously the Director of the Advanced Computing Research Centre (ACRC) at the University of Bristol, where he oversaw HPC, research data storage, and research software engineering services for the university. He also held the role of HPC Lead at Imperial College London.
Currently, Simon Burbidge leads a group of Research Software Engineers (RSEs) working on DiRAC HPC projects, focusing on benchmarking, performance monitoring, energy management, data storage, and software development in preparation for future exascale and next-generation computing technologies.
Simon has a strong background in collaborating with vendors and research communities, including ARM technologies, interconnects, and software development. He is active in HPC community and standards groups such as HPC-SIG and CIUK and is a member of the HPC User Forum Steering Committee. His work extends to advocating for diversity in HPC and developing training and education programs.
In summary, Simon Burbidge is a key figure in UK HPC infrastructure development and research computing leadership with significant contributions at the University of Bristol and beyond.hpcuserforum+5