The posting
Huawei Heisenberg Research Center (Munich) is responsible for advanced technology research, architectural development, design and strategic engineering of our products.
The Optical Technology Laboratory in Munich researches and develops signal processing, components, systems, and optical switching for fixed access, data center networks, metro and long-haul optical networks. The department engages in local research and collaborates with academic and industrial partners from the European ecosystem.
Datacenters are undergoing a paradigm shift driven by the explosion of AI/ML workloads, distributed cloud computing, and high-performance computing (HPC). Traditional electronic packet switching (EPS) faces severe scaling bottlenecks in bandwidth, power consumption, latency, and cost, while optical circuit switching (OCS) offers a promising alternative with high bandwidth density and lower energy per bit.
We are seeking a highly motivated PhD student to join us, focused on multi-tier OCS datacenter networks for AI infrastructure. The project proceeds in three phases: characterizing switching technologies across the nanosecond, microsecond, and millisecond regimes and their port-count trade-offs; architecting multi-tier fabrics that map these classes onto the scale-up, scale-in, scale-out, and scale-across tiers of the hierarchy (intra-node, rack/super-node, cluster, and cross-datacenter domains, respectively); and jointly optimizing traffic placement, scheduling, and reconfiguration so that switching time, power consumption, cost, and application performance (e.g., collective completion time and tail latency) are balanced for large-scale AI training and inference.
Join us as a
PhD Student — Multi-Tier Optical Circuit Switching for AI Datacenters (m/f/d)
Your mission
- Survey OCS switching technologies across the speed ladder — nanosecond-class burst and packet switching (e.g., semiconductor optical amplifiers, photonic integrated circuits), microsecond-class switches (e.g., piezo-actuated and fast MEMS, rotor switches, arrayed waveguide gratings), and millisecond-class space-division switches (e.g., 3D MEMS, liquid crystal, wavelength-selective switches) — and systematize their trade-offs in switching time, port count, insertion loss, power, reliability, and scalability.
- Build technology-level cost and power models covering switch fabrics, transceivers, lasers, amplifiers, fiber, and control overhead, enabling fair comparison of OCS classes on capital cost, operational power, and cost per switched bit at a given radix and reconfiguration time
- Develop analytical and simulation models of multi-tier OCS fabrics that capture reconfiguration delay, circuit setup overhead, congestion, and available bandwidth under demand-aware and demand-oblivious control, and validate them against published measurements and testbed experiments
- Formulate the tier-assignment and traffic-placement problem as a multi-objective optimization (e.g., mixed-integer programming, network flow, combinatorial algorithms) that maps parallelism dimensions and traffic classes — tensor-, pipeline-, data-, and expert-parallel flows — onto the appropriate OCS tier and port scale under performance, power, and cost constraints.
- Design scheduling and traffic-engineering algorithms that exploit mixed switching granularities, for example placing latency-sensitive collective communication on fast, small-radix tiers while serving elastic bulk traffic on slow, high-radix tiers, and reconfiguring a tier only when the amortized benefit exceeds its switching cost
- Evaluate the proposed architectures and algorithms on representative AI workloads, including MoE inference and super-node-scale training, through flow- and packet-level simulation, trace-driven analysis, and control-plane prototyping; publish results in top-tier venues such as JLT, JSAC, OFC, ECOC, SIGCOMM, and INFOCOM, and contribute to the doctoral thesis
Your areas of expertise
- A Master's degree (or equivalent) in Computer Science, Computer Engineering, Electrical or Optical Engineering, Applied Mathematics, Physics, or a related field, with an excellent academic record.
- Strong theoretical background in graph theory, combinatorial optimization, linear/integer programming, network flow, or queuing theory; experience with multi-objective optimization, and familiarity with learning-based methods (e.g., reinforcement learning, graph neural networks), is a plus
- Solid understanding of datacenter network architectures and high-performance interconnects (InfiniBand, RoCE, Ethernet, NVLink), and of optical switching technologies (OCS, WSS, ROADM) including their key metrics: switching speed, port count, insertion loss, and crosstalk
- Understanding of distributed deep learning workloads: collective communication primitives, data/tensor/pipeline/expert parallelism, and the traffic characteristics of large language model training and inference
- Strong programming skills in Python; experience with C++ or another systems language, mathematical solvers (e.g., Gurobi, CPLEX), and network or discrete-event simulators is expected
- A demonstrated ability to conduct research independently, evidenced by publications in peer-reviewed venues, a strong thesis, or comparable project work
- Excellent written and verbal communication skills in English and the ability to work in an international, interdisciplinary team
By applying to this position, you agree with our Recruitment Privacy Statement. You can read in full our privacy policy here.
Your rewards of working here
- Our culture is characterized by innovative power and team spirit as well as the intensive exchange of knowledge and experience within our global network.
- We offer healthy meals ranging from traditional Chinese to western delicacies in our famous company canteen.
- To keep your development ongoing, you will find a broad range of training opportunities. Many online and face-to-face training programs incl. language courses in German and Mandarin.
- Our diverse and welcoming environment is shaped by different backgrounds and around 40 individual nationalities.
- Self-responsible work in a competent, motivated and constantly growing team.
Please send your application and CV (incl. cover letter and reference letters) in English.
Huawei is a leading global information and communications technology (ICT) solutions provider. Our ICT solutions, products and services are used in more than 170 countries and regions, serving over one-third of the world's population. With 208,000 employees, Huawei is committed to develop the future information society and build a Better Connected World.
