Principal AI Performance & Reliability Engineer

AMD; San Jose (CA); Hybrid; USD 180,000 - 300,000; Full time

Job summary

AMD is seeking a Principal or Fellow level software engineer for the AI Infrastructure team in San Jose, CA or Bellevue, WA. You will optimize AI workloads across model training and inference, collaborating with customers and internal teams to identify bottlenecks and improve performance and reliability.

This role spans AI software and hardware stacks, requiring deep knowledge of architectures, ML frameworks, and production systems.

Qualifications

• PhD or master's degree in AI/ML/CS or related field.
• Strong expertise in AI infrastructure performance.
• Experience with production ML workloads.
• Excellent collaboration with customers and teams.

Responsibilities

• Profile and optimize AI training and inference workloads.
• Improve throughput, latency, memory efficiency and scalability.
• Identify bottlenecks across models, frameworks, and hardware.
• Develop performance tooling, benchmarks, and observability.
• Collaborate with ML engineers, systems engineers, and product teams.
• Document findings and recommendations.

Skills

Python

C++

PyTorch

TensorFlow

JAX

Profiling

Distributed systems

Computer architecture

Customer collaboration

Education

PhD or master's in AI/CS or related

Tools

ROCm

HIP

CUDA

Triton

XLA

MLIR

NCCL

Job description

AMD is seeking a Principal or Fellow level software engineer for the AI Infrastructure team in San Jose, CA or Bellevue, WA. You will optimize AI workloads across model training and inference, collaborating with customers and internal teams to identify bottlenecks and improve performance and reliability.

This role spans AI software and hardware stacks, requiring deep knowledge of architectures, ML frameworks, and production systems.