Senior GPU Inference Performance Engineer
Core
Own end-to-end performance analysis of GPU-accelerated AI inference workloads across the full stack from silicon to software runtime, driving competitive positioning against other accelerator vendors.
Role type
Senior IC GPU performance engineer (inference & distributed systems)
Builds
Competitive performance data, reproducible benchmarking harnesses, profiling scripts, and performance regression dashboards
Domain
AI infrastructure, GPU silicon, high-performance networking, distributed systems
Deliverable
production ML models
Required skills
GPU profiling (ROCm, CUDA, Nsight), distributed systems debugging, Linux systems analysis, RDMA/RoCE networking, Kubernetes GPU scheduling, Python, C/C++, GPU architecture knowledge
Preferred skills
HPC background, multi-node inference optimization, open-source infrastructure contributions, hardware/software evidence presentation
Technologies
ROCm, CUDA, Nsight, NCCL, RCCL, RDMA, RoCE, Kubernetes, Docker, containerd, HIP, MPI, UCX, ConnectX, Pensando
Responsibilities
Profile and diagnose bottlenecks in GPU runtimes, OS, drivers, and memory subsystems; execute head-to-head benchmarks between AMD and NVIDIA GPUs; optimize distributed inference topologies and network-level performance; build automation for trace collection and performance validation; analyze GPU operator and Kubernetes stack overhead
Seniority
Senior, hands-on IC
