Senior GPU AI Inference Performance Engineer

Advanced Micro Devices, Inc.; Santa Clara (CA); On-site; USD 180,000 - 280,000; Full time

Job summary

Advanced Micro Devices, Inc. (AMD) in Santa Clara, CA is seeking a Senior GPU Inference Performance Engineer to own end‑to‑end performance analysis of GPU‑accelerated AI inference workloads. You will profile hardware, software runtimes, and network interactions to explain bottlenecks with hardware‑rooted evidence.

You will drive cross‑stack profiling, benchmarking, and optimization across AMD and NVIDIA ecosystems, including multi‑server inference and Kubernetes integration.

Qualifications

• Background in GPU performance engineering, HPC, distributed systems, networking, or systems performance analysis.
• Hands-on proficiency with ROCm/NVIDIA profiling toolchains and deep GPU architecture knowledge.
• Experience with Linux performance analysis, containers, and kernel-level software.

Responsibilities

• Full‑stack GPU profiling across AMD and NVIDIA GPUs to identify bottlenecks.
• Profile interactions between GPU runtimes, Linux, drivers, and containers to improve throughput.
• Design and run head‑to‑head benchmarks on AI workloads with data‑backed explanations.
• Profile multi‑server inference networking and measure impact of latency and bandwidth.
• Profile Kubernetes GPU scheduling and container runtimes to reduce latency and jitter.
• Build reproducible benchmarking harnesses and profiling scripts.

Skills

GPU performance engineering

Python

C/C++

Linux system performance analysis

Kubernetes GPU scheduling

Education

Bachelor's degree in Computer Science, Computer Engineering, Electrical Engineering, or related field

Tools

ROCm toolchain

NVIDIA Nsight Systems/Compute

NCCL/RCCL

GPUDirect RDMA

RDMA/RoCE

UCX

MPI

Job description

Advanced Micro Devices, Inc. (AMD) in Santa Clara, CA is seeking a Senior GPU Inference Performance Engineer to own end‑to‑end performance analysis of GPU‑accelerated AI inference workloads. You will profile hardware, software runtimes, and network interactions to explain bottlenecks with hardware‑rooted evidence.

You will drive cross‑stack profiling, benchmarking, and optimization across AMD and NVIDIA ecosystems, including multi‑server inference and Kubernetes integration.