Distinguished Technologist - AI Model Performance Architect
Core
Bridging system memory architecture with AI model behavior to ensure optimal performance through HW/SW co-design and workload-aware tuning.
Role type
Distinguished Technologist (AI Model Performance Architect)
Builds
AI workload optimization strategies for faster model startup, improved inferencing efficiency, and reduced memory needs.
Domain
Electrical hardware design, AI systems, and memory architecture.
Deliverable
production ML models
Required skills
Memory & AI model design, Computer Science, Electrical Engineering, Hardware Architecture, FPGA, Systems Design, MATLAB, Simulations, Debugging, Hardware Design
Preferred skills
New Product Development, Physics, Mechanical Engineering, Schematic Capture, Oscilloscope, Printed Circuit Board
Technologies
FPGA, MATLAB, Hardware Architecture
Responsibilities
Analyze system bandwidth, latency, and compute bottlenecks; Model performance across CPU/GPU/NPU for AI workloads; Simulate workload execution (PnP tradeoffs, resource contention); Optimize model execution (initialization time, caching, KV reuse); Drive HW/SW co-design (memory, caching, scheduling); Partner with ML teams to align model structure with system architecture; Develop technology roadmaps for electrical/hardware engineering teams; Provide mentoring and guidance to less-experienced staff members.
Seniority
Distinguished, strategy & mentorship
