Initiative overview
AIRR addresses a shortage of publicly available computing resources in the UK's research and development community. It is positioned as a critical enabler of AI innovation and the government's broader growth agenda, providing access to advanced hardware, software, and expert support through AI-optimised high-performance computing platforms. Use cases span growth sectors including healthcare and clean energy, as evidenced by AIRR Phase 1 projects.
The programme's two flagship clusters are Isambard-AI at the University of Bristol , the UK's most powerful public compute facility, comprising 5,448 Nvidia GH200 Grace-Hopper superchips supplied by HPE, and Dawn at the University of Cambridge, made up of 1,024 Intel Data Centre GPU Max 1550 GPUs, operated in close partnership with Intel and Dell at the Cambridge Open Zettascale Lab. The government has committed over £350 million to these two clusters by 2030.
Access is structured around three routes managed through the AIRRPortal. Rapid Access targets UK-registered micro, small and medium organisations for early-stage AI product development, offering up to 20,000 GPU hours within three months. Gateway supports first-time users from academia or industry for algorithm testing and benchmarking, also providing up to 10,000 GPU hours within three months. The Innovator route is designed for larger research partnerships, offering between 50,000 and 150,000 GPU hours over six months for activities including novel algorithm development and AI-driven data synthesis. All projects must be conducted in the UK, aimed at delivering UK benefits, and access is governed by the Subsidy Control Act 2022.
