AMD announced a commitment of up to £2 billion over five years to strengthen the United Kingdom's artificial intelligence infrastructure, scientific research capabilities, and technical workforce. Announced by AMD Chair and CEO Dr. Lisa Su at London Tech Week, the investment spans new strategic partnerships with universities and research organizations, support for national supercomputing projects, and alignment with the UK government's AI Opportunities Action Plan and AI Hardware Strategy.
The UK government is actively building out what it calls sovereign AI infrastructure domestically owned and operated compute capacity that reduces dependence on foreign cloud providers for critical national research and public-sector AI workloads. AMD's investment directly supports this national priority, and the announcement was publicly welcomed by both the Chancellor of the Exchequer and the Technology Secretary signaling strong government alignment with the initiative.
For AMD, the UK represents a strategically important market with world-class universities, a strong research culture, and growing public and private investment in AI infrastructure. Establishing deep partnerships now positions AMD's hardware and software stack particularly its Instinct GPUs, EPYC CPUs, and ROCm open software as foundational to the UK's AI ecosystem for the coming decade.
Imperial College London: AMD is collaborating with Imperial on computational science research, including healthcare innovation and climate modeling. The partnership will also explore optimizing AI models and scientific workflows specifically for AMD's compute platforms and ROCm software stack. This is notable because ROCm is AMD's open-source alternative to NVIDIA's CUDA, and embedding it into leading academic research workflows builds the developer ecosystem AMD needs to compete more broadly in AI infrastructure.
Oriole Networks and the ARIA Scaling Inference Lab: AMD is supporting the UK's Advanced Research and Invention Agency initiative to address AI infrastructure bottlenecks. The collaboration pairs Oriole's PRISM photonic networking architecture with AMD Instinct GPUs and EPYC processors. Photonic networking uses light rather than electrical signals to move data between compute nodes, offering dramatically lower latency and better energy efficiency at scale. This initiative is expected to produce the world's first large-scale AI system running on a pure photonic network a potentially significant step forward for next-generation AI infrastructure architecture.
University of Cambridge, Zenith and Sunrise Supercomputers , AMD and Dell Technologies are jointly supporting two major AI supercomputing systems at Cambridge:
• Zenith is a new national AI-for-science platform funded by the UK Department for Science, Innovation and Technology and UK Research and Innovation. It is designed and operated by Cambridge, built on AMD and Dell technology, and intended to support a broad range of scientific AI applications.
• Sunrise is a second supercomputer currently under construction, funded by the Department for Energy Security and Net Zero, owned by the UK Atomic Energy Authority, and operated by Cambridge. Sunrise is dedicated specifically to fusion energy research supporting the UKAEA's long-term mission to develop commercially viable nuclear fusion. Combining AI with fusion research allows scientists to model plasma behavior, optimize reactor designs, and accelerate the experimental cycle in ways that were previously computationally impractical.
Together, Zenith and Sunrise cover an impressive breadth of national research priorities healthcare, climate modeling, materials science, engineering simulation, fusion energy, and scientific AI model development.
Across all these initiatives, AMD is deploying a consistent hardware and software platform:
• AMD Instinct GPUs the compute backbone for AI training and inference workloads
• AMD EPYC CPUs high-core-count server processors for data-intensive scientific workloads
• AMD ROCm open software the open-source GPU compute platform, positioned as a more flexible and vendor-neutral alternative to proprietary GPU software ecosystems
The consistent use of ROCm across academic and research deployments is strategically important it builds familiarity and optimization expertise in the UK research community that benefits AMD's commercial positioning over time.
