What are the responsibilities and job description for the HPC Infrastructure & Cluster Engineer- TS/SCI Cleared position at Marathon TS?
HPC Infrastructure & Cluster Engineer
Springfield, VA
Marathon TS is seeking a TS/SCI-cleared Infrastructure Engineer to manage and optimize a high-performance compute environment supporting intensive AI/ML workloads. This role is heavily focused on Linux cluster administration, GPU infrastructure, workload scheduling, high-speed networking, storage, and containerized environments.
Key Responsibilities:
- Cluster Administration: Manage the day-to-day operations of the customer compute cluster, including Linux operating system administration, hardware monitoring, patching, and system upgrades.
- Resource and Job Management: Configure, maintain, and optimize workload management and orchestration platforms, utilizing the Run:AI job scheduler to ensure efficient distribution of intensive AI/ML workloads across the cluster.
- Infrastructure Optimization: Tune cluster performance at the hardware, operating system, and network levels to maximize compute efficiency and data throughput for customer workloads.
- Storage and Network Management: Administer storage solutions and high-speed networking fabrics. Support the transition to and ongoing management of an InfiniBand GPU-to-GPU network infrastructure to minimize latency for distributed operations.
- Environment Configuration: Partner with technology integration teams to provision specific environments, dependencies, and container platforms, specifically leveraging Red Hat OpenShift, required for seamless customer model deployment.
- Security and Compliance: Ensure all infrastructure components remain compliant with federal security standards, implementing strict access controls and maintaining system accreditations.
Basic Qualifications:
- Clearance: Active TS/SCI with the ability to obtain CI Poly.
- Experience: 5 years of experience in Linux systems administration and infrastructure management with a specific focus on high-performance computing environments.
- Technical Skills:
- Expertise in managing bare-metal servers, enterprise storage arrays, and advanced network configurations (Experience with InfiniBand).
- Strong proficiency with workload managers, job schedulers, and AI orchestration tools (e.g., Run:AI, SLURM).
- Hands-on experience with enterprise container orchestration platforms, specifically OpenShift or Kubernetes.
- Experience writing automation and configuration scripts (e.g., Bash, Python) to streamline cluster maintenance.
- Troubleshooting Focus: Proven ability to diagnose and resolve complex hardware, network, and OS-level issues.
Preferred Qualifications:
- Familiarity with parallel file systems and high-throughput storage architectures.
- Prior experience engineering or managing high-speed GPU-to-GPU communication topologies.
US Citizenship Required
TS/SCI Required