Senior Network Systems Researcher
Join an accelerated-computing engineering organization researching Ethernet networks, switch ASICs, and high-performance AI fabrics. The role focuses on connecting switch-level capabilities to cluster-scale outcomes and influencing next-generation network architecture.
Responsibilities
- Research how switch-level capabilities affect cluster-scale outcomes, including orchestration, job scheduling, collective communication, and observability
- Evaluate hardware and software trade-offs for high-radix Ethernet switches and next-generation AI fabrics
- Investigate machine-learning methods for traffic analysis, control-plane assistance, and configuration optimization
- Build proof-of-concept implementations using network simulations, performance models, and data-driven analysis
- Assess emerging protocols, software stacks, open-source projects, partner ecosystems, and academic research against the Ethernet switch roadmap
- Collaborate with architects, advanced development teams, and researchers to shape networking solutions and architecture decisions
Requirements
- M.Sc. or Ph.D. in Computer Science, Computer Engineering, or Electrical Engineering
- At least 12 years of experience in system architecture, networking technologies, and embedded software, including at least 3 years in research
- Expert knowledge of Ethernet and IP technologies, network topologies, and data-center networking protocols and features
- Strong understanding of large-scale network behavior, network simulation, and distributed-computing workload effects on networks
- Demonstrated ability to conduct research, prototype concepts, validate their value, and develop practical applications
- Strong written and verbal communication skills, including consensus-building in large organizations
- Strong problem-solving and critical-thinking abilities
Nice to have
- Practical networking research and innovation experience
- Knowledge of switch ASIC hardware and software abstractions
- Experience applying machine learning or artificial intelligence to networking problems