Postdoctoral Researcher – System-Level Evaluation Framework for AI Architectures

About the role

The Electronics group at the Lyon Institute of Nanotechnology (INL) is looking for a postdoctoral researcher to create an open, modular and extensible system‑level evaluation framework for emerging AI computing architectures. The framework will support design‑space exploration and co‑optimization of technology, architecture and workloads, linking device‑level results to system‑level performance metrics.

Key responsibilities

• Develop a parameterised architectural modelling environment that transforms low‑level device, circuit and array data into system‑level performance estimates.
• Implement both bottom‑up projection from technology choices to system metrics and top‑down constraint propagation from energy, latency and accuracy targets to technology requirements.
• Integrate and extend existing INL tools for predictive system assessment, emerging memories, compute‑in‑memory and photonic computing, ensuring openness and reproducibility.
• Provide cross‑layer analyses that expose bottlenecks, sensitivities, feasibility limits and break‑even points for emerging AI accelerators.

Required profile

• PhD in Electrical/Electronic Engineering, Computer Engineering, Computer Science or a related discipline.
• Strong background in computer architecture, AI accelerators, performance/energy/area modelling, design‑space exploration and multi‑objective optimisation.
• Experience with hardware‑software co‑design, workload mapping and emerging memory or photonic computing technologies.

Required skills

• Python programming and solid software‑engineering practices.
• SystemC and C/C++ development.
• RISCV simulation and QEMU.
• PyTorch and ONNX workflow handling.
• Familiarity with architecture simulators such as NeuroSim, CiMLoop/AccelForge or Timeloop.