Lead - Soc Architect

AMD AMD · Semiconductors · San Jose, CA · Engineering

Lead SoC Architect role focused on defining the architecture for next-generation physical AI SoCs, including AI acceleration, memory systems, and interconnects. The role involves system-level innovation, cross-functional guidance, and customer collaboration, with a strong emphasis on performance, power, and cost trade-offs for AI and other emerging workloads.

What you'd actually do

  1. Own and lead the architecture for next‑gen physical AI SoCs, spanning processor subsystems, GPU/AI acceleration, coherent and non‑coherent interconnects, memory hierarchy, and advanced I/O.
  2. Define long‑term SoC architectural strategy aligned with product, roadmap, and customer requirements.
  3. Drive system‑level analysis, including performance modeling, power modeling, silicon feasibility studies, and early floorplan architecture.
  4. Develop clear, detailed architecture specifications for compute, memory, interconnect, safety/security, boot/configuration, and SoC management subsystems.
  5. Guide cross‑functional execution, providing leadership across RTL design, verification, physical design, firmware, software, and platform teams.

Skills

Required

  • SoC architecture
  • heterogeneous compute
  • AI acceleration
  • memory systems
  • interconnects
  • performance modeling
  • power modeling
  • RTL design
  • verification
  • physical design
  • firmware
  • software

Nice to have

  • x86 and ARM based processor subsystems
  • GPU compute architecture
  • AI accelerators
  • DSP pipelines
  • interconnect fabrics
  • memory technology
  • Automotive
  • Wireless systems
  • AI/ML use-cases
  • high-speed I/Os
  • multi-chiplet partitioning
  • chiplet interconnect standards
  • architectural modeling
  • performance simulation frameworks
  • ecosystem partners
  • customer roadmaps

What the JD emphasized

  • recognized industry expert
  • deep experience
  • track record of delivering differentiated SoC solutions
  • patents, publications, or major architectural wins strongly preferred

Other signals

  • AI acceleration
  • AI inference/training
  • heterogeneous compute
  • SoC architecture