Performance Modeling Engineer
Core
Develop performance models and projections for hardware architectures to optimize throughput and latency for deep learning inference workloads.
Role type
Performance Modeling Engineer (Hardware/ML Systems)
Builds
Hardware architectures for frontier AI inference systems
Domain
Computer Architecture / Deep Learning Infrastructure
Deliverable
production ML models
Required skills
performance modeling, micro-architecture analysis, deep learning workload profiling, hardware/software co-optimization, regression testing, system design pathfinding
Preferred skills
GPU architecture knowledge (CUDA), multi-chip inference mapping, transformer model optimization, architecture simulators (gem5), ASIC/FPGA development, published research in computer architecture
Technologies
CUDA, gem5, trace-driven simulators, GPUs, TPUs, ASICs, FPGAs
Responsibilities
Develop performance models across varying workloads and configurations; Profile and analyze deep learning workloads to identify bottlenecks; Drive hardware/software co-optimization; Run regressions to validate models against real systems; Inform next-generation architectural decisions
Seniority
Mid-Senior, hands-on IC
