Principal Silicon Performance Architect
Core
Extend and adapt simulation infrastructure to model new micro-architecture innovations for AI inference, analyzing performance across latency, throughput, and efficiency.
Role type
Principal Silicon Performance Architect
Builds
AI inference simulation infrastructure and performance models
Domain
Semiconductor chip architecture and AI inference acceleration
Deliverable
production ML models
Required skills
C/C++ programming, Python programming, micro-architecture analysis, performance modeling, simulation tool modification, profiling, bottleneck analysis, experimental design, AI inference acceleration analysis, compiler and runtime stack knowledge, GPU performance analysis
Preferred skills
Advanced degree in Electrical or Computer Engineering, experience with large-scale experiment generation, AI-assisted workflow design
Technologies
C, C++, Python, GPU, compilers, runtimes, model serving systems
Responsibilities
Analyze performance for current and forward-looking AI inference workloads, drive design-space exploration using AI-assisted workflows, collaborate with chip, system, and software architects to evaluate architectural variations, modify architectural simulators or performance modeling codebases
Seniority
Principal, hands-on IC with strategy influence
