Senior NoC Research Engineer – Network-on-Chip

Posted 4hrs ago

Employment Information

Education
Salary
Experience
Job Type

Report this job

Job expired or something wrong with this job?

Job Description

Senior NoC Research Engineer shaping interconnect architecture for SpiNNcloud’s brain-inspired computing platform. Modeling ML workloads and optimizing chip-scale data movement, performance, and energy efficiency.

Responsibilities:

  • Explore and evaluate NoC topologies, router and switch designs, arbitration, flow control, QoS and congestion control
  • Build and maintain architectural models and simulations using SST, gem5 or custom cycle-level and analytical models
  • Develop traffic generators reflecting real ML workloads, including collective communication, tensor parallelism and memory traffic
  • Analyse power, performance and area (PPA) trade-offs and support recommendations with quantitative evidence
  • Ensure deadlock- and livelock-free designs through analysis, formal methods or simulation
  • Investigate adaptive routing strategies and topology options
  • Translate complex protocol and IP specifications, including AXI, CHI and APB, into clear design requirements
  • Evaluate third-party interconnect IP and support build-versus-buy decisions
  • Assess how RDMA concepts and chip-to-chip interfaces shape data movement between accelerators at system scale
  • Present findings and recommendations to architecture and leadership teams
  • Contribute to publications and patents in on-chip networks and data movement for ML
  • Report to the CTO and collaborate closely with chip architecture, hardware and software teams

Requirements:

  • PhD, or Master's with equivalent research experience, in Electrical Engineering, Computer Engineering, Computer Science or a related field
  • Senior-level track record in NoC or interconnect architecture, in industry or applied research
  • Strong background in on-chip interconnect and traffic management: NoC topologies, routers and switches, arbitration, flow control, QoS and congestion control
  • Good understanding of ML workloads and accelerator architectures, including collective communication, tensor parallelism and memory traffic
  • Working knowledge of on-chip protocols such as AXI, CHI or AHB
  • Understanding of RDMA concepts and transports (InfiniBand, RoCE), including queue pairs, memory registration and zero-copy transfers
  • Experience reasoning about PPA trade-offs with quantitative evidence
  • Hands-on architectural modelling and simulation using SST, gem5 or own models
  • Strong Python and C++ skills
  • Ability to read complex protocol and IP specifications and turn them into clear design requirements
  • Experience with deadlock and livelock avoidance, adaptive routing or topology exploration
  • Clear written and verbal communication in English
  • Proactive, collaborative approach to open-ended problems
  • Experience with chip-to-chip and die-to-die interfaces such as PCIe, Ethernet or UCIe is preferred
  • Experience with memory subsystems such as HBM and CXL is preferred
  • Publications or patents in on-chip networks, interconnect architecture, or data movement for ML or HPC are preferred
  • Familiarity with massively parallel architectures is preferred
  • Applicants must provide their current location and state whether visa sponsorship and/or relocation support is needed to work in Germany

Benefits:

  • 30 days annual vacation
  • Up to 2 weeks of work abroad
  • Work on cutting-edge sparse computing technology
  • High-impact role with direct influence on the architecture of SpiNNext
  • Collaborative team culture with experienced engineers from academia and industry
  • Equity participation in a growing deep-tech company