Member of Technical Staff (Hardware)
Core
Design and run performance benchmarks across AI accelerators (GPUs, TPUs, custom silicon) to measure throughput, latency, and price-performance, while building evaluation methodologies and cost models.
Role type
Senior IC hardware benchmarking engineer (AI accelerators)
Builds
Hardware benchmarking stack (AA-AgentPerf), performance analysis reports, and cost/throughput models for the AI industry
Domain
AI hardware / Silicon / Inference infrastructure
Deliverable
production ML models
Required skills
Python, data analysis, CUDA, TensorRT, vLLM, inference economics, AI accelerator architecture
Preferred skills
Experience with frontier AI models, relationship management with chipmakers
Technologies
CUDA, TensorRT, vLLM, Python
Responsibilities
Design and execute performance benchmarking across GPUs, TPUs, and custom inference silicon; Develop and maintain frameworks for measuring AI hardware performance and efficiency; Analyze inference economics including cost-per-token and utilization; Partner with top hardware companies to benchmark latest silicon; Produce reports and data visualizations on hardware performance and economics
Seniority
Senior, hands-on IC