Principal System Performance Architect
Core
Define performance telemetry, profiling, and optimization architecture for next-generation AI infrastructure, edge platforms, and physical AI systems.
Role type
Principal System Performance Architect
Builds
System-wide performance observability, profiling capabilities, and analysis strategies for heterogeneous compute platforms.
Domain
Semiconductor / System Architecture / AI Infrastructure
Deliverable
production ML models
Required skills
System architecture, performance engineering, profiling methodologies, heterogeneous computing systems, hardware performance counters, trace mechanisms, power-performance optimization, bottleneck analysis, workload characterization, cross-stack data correlation (hardware to software), architectural tradeoff analysis.
Preferred skills
AI infrastructure platforms, distributed AI systems, edge AI, robotics, automotive compute, PCIe/CXL/Ethernet fabrics, RDMA, platform power management, thermal behavior, cloud-scale observability.
Technologies
Linux perf, NVIDIA Nsight, Intel VTune, Arm Performance Studio, Arm Streamline, Windows Performance Analyzer, PCIe, CXL, Ethernet, RDMA, NVLink, UALink.
Responsibilities
Define telemetry architecture for CPUs, accelerators, memory, interconnects, and I/O; drive platform specifications for performance observability; collaborate with SoC, silicon, and power architects to embed profiling capabilities; develop analysis strategies for heterogeneous systems; translate requirements into scalable telemetry models; work with tooling teams to expose architectural hooks for analysis environments; participate in silicon bring-up, debug, and validation activities.
Seniority
Principal, hands-on IC with strategic guidance
