Senior Researcher AI Computer Architecture

Salary: £70,000 - 100,000 per year

Requirements:

• We require a PhD in electronic engineering, computer architecture, computer systems, or a related field, or equivalent training and research experience.
• We require at least 3 years of related research or equivalent industrial experience.
• We require demonstrated architectural impact and experience in computer system organisation.
• We require a proven record of working at the interface of different research fields or in multidisciplinary teams.
• We require the ability to work effectively in a multidisciplinary environment across hardware, systems, networking, and AI workloads.
• We require excellent written and spoken English communication skills, including the ability to clearly present technical results and justify assumptions to diverse technical audiences.
• We prefer a strong understanding of techniques, terminology, and frameworks for modern large-scale machine learning systems.
• We prefer experience in research or industry projects related to AI systems design, architecture modelling, and/or simulation.
• We prefer experience in scale-up networking in AI GPU/xPU systems.
• We prefer experience in memory systems for AI inference and training infrastructure.
• We prefer experience building high-performance LLM inference systems using SGLang or vLLM.
• We prefer publications in top computer architecture, systems, and/or ML conferences.

Responsibilities:

• We define the research roadmap and advance a long-term agenda for AI system architectures built on novel optical interconnects, new memory technologies, and other emerging hardware trends.
• We identify research gaps across optical IOs, memory systems, and AI workloads, and lead influential projects in these areas.
• We work closely with experts in adjacent disciplines to steer the overall research programme and align on the most impactful strategic priorities.
• We collaborate across domains and stakeholders to identify and solve problems in a multidisciplinary team.
• We design, model, and prototype future AI systems through an integrated co-design approach.
• We build system models and simulations to explore trade-offs in scale-up networking, memory offloading, and memory disaggregation.
• We characterise agentic AI workload demands and develop software prototypes to validate concepts end to end.
• We plan and execute experiments, collect data, analyse results, and validate or disprove hypotheses to guide research impact.
• We collaborate with external industry partners to drive hardware proof-of-concept development and production roadmaps.
• We expand relationships with hardware vendors, research teams, and production teams inside and outside Microsoft.
• We write, create, and deliver technical presentations for internal and external stakeholders to share research outcomes and strategic insights.
• We mentor and grow team capabilities by coaching research interns and junior researchers, and by fostering shared expertise across compute, networking, and memory systems.

Technologies:

• Agentic AI
• AI
• Hardware
• LESS
• LLM
• Machine Learning
• iOS
• vLLM
• Architect
• Network

More:

We are Microsoft Research Cambridge, and our Future AI Infrastructure (FAI) team is pioneering new hardware and system technologies to reshape how AI systems scale by overcoming bandwidth, latency, and energy constraints in todays accelerator architectures. This is a full-time individual contributor role based in Cambridge, United Kingdom, with a hybrid schedule of 3 days per week in the office and travel of less than 25%. The role sits within Research Sciences at IC4 level and offers a typical UK base pay range of 74,700.00 to 122,600.00 per year, with eligibility for benefits and other compensation depending on the role. We are an inclusive employer and are committed to a culture of respect, integrity, and accountability.

last updated 34 week of 2026