Senior Cluster Reliability Architect - AI/HPC Platforms

Socket.dev; Austin (TX); On-site; USD 180,000 - 260,000; Full time

Benefits offered by this job

AMD Benefits

Job summary

AMD is seeking a Principal Cluster Reliability Architect to define and drive the reliability strategy for next-generation AI and HPC cluster platforms. This role serves as the technical authority for end-to-end cluster reliability, spanning compute, networking, storage, and control plane architectures.

The successful candidate will influence architecture, deployment, validation, and operational readiness across AMD reference designs and large-scale customer deployments.

Qualifications

• Bachelor's degree in Computer Engineering, Computer Science, Electrical Engineering, or a related technical discipline.

Responsibilities

• Define and implement a comprehensive cluster reliability architecture framework across the full infrastructure lifecycle.
• Establish reliability requirements for reference architectures and customer deployments.
• Drive architecture decisions to improve fault tolerance, redundancy, and recoverability.
• Partner with SRE and Platform Operations to define sustainable operational models.
• Define standards for observability, health monitoring, and lifecycle management.

Skills

Cluster architecture

Distributed systems

Linux systems

Production-scale infra

Kubernetes

Slurm

ROCm

GPU clusters

AI training

Inference infra

Education

Bachelor's degree in Computer Engineering, Computer Science, Electrical Engineering

Tools

InfiniBand

RoCE

Job description

AMD is seeking a Principal Cluster Reliability Architect to define and drive the reliability strategy for next-generation AI and HPC cluster platforms. This role serves as the technical authority for end-to-end cluster reliability, spanning compute, networking, storage, and control plane architectures.

The successful candidate will influence architecture, deployment, validation, and operational readiness across AMD reference designs and large-scale customer deployments.