AI Infrastructure Engineer - Slurm Platform

Advanced Micro Devices, Inc.•San Jose, CA

60d•Hybrid

About The Position

At AMD, our mission is to build great products that accelerate next-generation computing experiences-from AI and data centers, to PCs, gaming and embedded systems. Grounded in a culture of innovation and collaboration, we believe real progress comes from bold ideas, human ingenuity and a shared passion to create something extraordinary. When you join AMD, you'll discover the real differentiator is our culture. We push the limits of innovation to solve the world's most important challenges-striving for execution excellence, while being direct, humble, collaborative, and inclusive of diverse perspectives. Join us as we shape the future of AI and beyond. Together, we advance your career. THE ROLE: A senior technical contributor that drives end-to-end delivery of software solutions, directly contributing to, and coordinating implementation and optimization across multiple teams for AI software development, inference and training of machine learning models. The position will involve interfacing with software and hardware engineering teams and AMD partners to plan, develop and optimize use cases. This is an exciting opportunity to work on the cutting edge of GPU Computing for Machine Learning. THE PERSON: We are seeking a DevOps Engineer / HPC Platform Engineer to build and operate our Slurm-based GPU compute platform, which complements our Kubernetes-based orchestration layer for GPU and CPU workloads. The ideal candidate should be passionate about software engineering and possess leadership skills to independently deliver on multi-quarter projects. They should be able to communicate effectively and work optimally with their peers within our larger organization. Finally, you aren't afraid of a team in more of a startup mode at a larger company and willing to jump in to help in areas adjacent to your main project as needed.

Requirements

8+ years of experience managing and automating HPC or Slurm clusters in production environments.
Deep understanding of Linux systems, job schedulers (Slurm), and resource management for GPU-accelerated workloads.
Strong troubleshooting skills across compute, storage, and network layers.
Proven ability to collaborate with developers and researchers to design scalable HPC solutions.

Nice To Haves

Experience integrating Slurm with Kubernetes or other control planes.
Experience with HPC storage and I/O technologies (Lustre, ZFS, WekaFS, NFS).
Familiarity with metrics collection and visualization using Prometheus, Grafana, and Thanos.
Exposure to CI/CD pipelines and DevOps practices for scientific or ML workloads.
Understanding of machine learning workflows and frameworks (PyTorch, vLLM, SGLang).
Experience with infrastructure automation (e.g., Ansible, Terraform) and scripting (Python, Bash).

Responsibilities

Design, deploy, and operate Slurm clusters across on-prem and multi-cloud GPU environments (Azure, OCI, Vultr, DigitalOcean, etc.).
Integrate Slurm with the broader orchestration ecosystem, enabling hybrid scheduling, unified authentication, and telemetry pipelines.
Build platform features that improve developer experience - e.g., job submission APIs, automated environment setup, and metrics dashboards.
Optimize cluster utilization and scheduling for GPU and CPU workloads; develop fair-share, QoS, and preemption policies.
Monitor cluster health and performance, implementing observability pipelines using Prometheus, Grafana, and custom exporters.
Collaborate with internal developers (framework, compiler, and application teams) to understand workload needs and translate them into scalable Slurm features.
Contribute to storage and network integration, ensuring performant I/O (e.g., NFS, Lustre, Weka) and high-speed interconnect configuration (InfiniBand, NVLink).
Support the job lifecycle - from image builds and environment modules to debugging and performance tuning of Slurm jobs.