Senior Performance Modeling Engineer
Core
Develop and enhance analytical performance models and simulation frameworks for AI accelerators, GPU architectures, and large-scale AI systems to optimize throughput, latency, and efficiency.
Role type
Senior IC performance modeling engineer (AI accelerators)
Builds
System-level performance models, design-space exploration tools, and visualization frameworks for accelerator and cluster configurations.
Domain
AI Systems / Computer Architecture
Deliverable
production ML models
Required skills
Analytical performance modeling, roofline modeling, architectural simulation, workload mapping, parallelization analysis (tensor/pipeline/expert/data/context), Python, C/C++, performance profiling, benchmarking, GPU programming, knowledge of transformers/MoE/attention mechanisms, TCO analysis.
Preferred skills
AI agentic systems & flow for performance modeling, PyTorch, vLLM, SGLang, distributed runtime systems, HW-SW co-design.
Technologies
Python, C/C++, PyTorch, vLLM, SGLang, GPU architectures, AI training/inference frameworks.
Responsibilities
Model end-to-end AI training and inference workloads including compute kernels and memory access patterns; Evaluate architectural features and interconnects to identify performance improvements; Perform workload mapping and parallelization analysis across various strategies; Correlate analytical model predictions with silicon measurements and software traces; Build and automate performance modeling and design-space exploration tools; Collaborate with hardware and software teams to translate insights into architecture recommendations.
Seniority
Senior, hands-on IC