Physical AI SoC Architect
Core
Define next-generation System-on-Chip (SoC) architectures for automotive, robotics, vision, and Physical AI applications by translating real-world workloads into quantitative requirements for compute, memory, bandwidth, and latency.
Role type
Senior IC SoC Architect (Physical AI & Automotive)
Builds
Next-generation heterogeneous SoCs for automotive, robotics, and vision systems
Domain
Semiconductor / Hardware Architecture / Physical AI
Deliverable
production ML models
Required skills
Heterogeneous SoC architecture, System performance analysis, AI/ML inference, Computer vision, DDR/LPDDR memory systems, NoC/interconnect traffic modeling, Python/C/C++/SystemC/MATLAB simulation, Data-flow diagramming, Power/performance/area/cost trade-offs
Preferred skills
x86/ARM/RISC-V processor architecture, Cache/memory hierarchy design, Functional safety, Real-time performance constraints, Embedded system constraints
Technologies
Python, C/C++, SystemC, MATLAB, DDR, LPDDR, NoC, x86, ARM, RISC-V
Responsibilities
Translate Physical AI and automotive workloads into SoC compute and memory requirements; Model end-to-end data flows across sensors, CPUs, AI accelerators, and memory hierarchies; Analyze and optimize data placement across DDR, caches, and on-chip memory; Perform what-if and scaling analyses for camera count, resolution, and AI model characteristics; Develop data-flow diagrams, bandwidth models, and latency budgets; Evaluate architectural alternatives for performance, power, area, and cost; Collaborate with hardware, software, and performance teams to optimize system performance.
Seniority
Senior, hands-on IC