Cambridge Residency Programme: Next-generation networks for AI Datacentre Networks

Overview Microsoft Research Cambridge, UK is seeking a postdoctoral researcher to advance fundamental and applied research on designing nextgeneration datacentre networks for AI workloads. As largescale AI training and inference continue to place unprecedented demands on network performance, energy efficiency, reliability, and cost, there is a growing need to rethink traditional packetswitched architectures and explore alternatives that better align with emerging communication patterns.

The position is based at Microsoft Research Cambridge within the Future AI Infrastructure group. The lab conducts world-leading research that has influenced many of Microsoft’s products and partnerships. The group is highly cross-disciplinary, spanning networks, hardware, optics, computer science and engineering, and works closely with product teams and academic partners to tackle the challenges of next generation cloud infrastructure.

This role offers the opportunity to conduct cuttingedge research at the intersection of network protocols, systems design, and AI infrastructure, with direct relevance to real-world systems at global scale. You will contribute new ideas, extend existing models and tools, validate them on experimental testbeds, and collaborate with researchers and engineers to explore their broader implications for future cloud platforms.

• Contract Duration: 2 Years
• Location:  Cambridge, UK

Responsibilities You will advance the research on AI networks through design, simulations, testbed cross-validation, and implementing promising solution into frameworks for demonstration.

• Evaluate the performance of novel network architecture for AI using analytical modelling and simulation
• Validate analytical models using our experimental testbed.
• Work closely with collaborators within our team, in product teams, and other research groups to expand our tools to solve larger challenges
• Expand existing tools and analysis to cover scenarios of interest to product teams.
• Design new network architectures leveraging new emerging hardware such as co-package optics, optical circuit switches, and new interconnects.

Qualifications Required/Minimum Qualifications :

• A PhD in a relevant field (e.g. ML, Computer Science, Computer Engineering, Mathematics) or equivalent industry experience.
• Strong understanding and hands-on experience with one or more of: analytical modelling of distributed computer systems (e.g. alpha-beta or flow-based models for networks), inference software stacks and related tools.
• Ability to clearly communicate your research to a diverse audience.
• Adaptability to engage in and embrace new research domains.

Preferred/Additional Qualifications :

• Strong coding (python and C++) and engineering skills.
• Experience with circuit switched network and/or modelling and implementing collectives.
• Experience with the deployment of large language models.
• Experience with CUDA/RDMA programming.

This position will be open for a minimum of 5 days, with applications accepted on an ongoing basis until the position is filled.

Microsoft is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to age, ancestry, citizenship, color, family or medical care leave, gender identity or expression, genetic information, immigration status, marital status, medical condition, national origin, physical or mental disability, political affiliation, protected veteran or military status, race, ethnicity, religion, sex (including pregnancy), sexual orientation, or any other characteristic protected by applicable local laws, regulations and ordinances. If you need assistance with religious accommodations and/or a reasonable accommodation due to a disability during the application process, read more about requesting accommodations.