HPC Infrastructure/Network Engineer

HPC Infrastructure/Network Engineer

Location: Ashburn, VA - Onsite

Duration: 6 to 8 months

Role Overview:

The HPC (High-Performance Computing) Role focuses on planning, implementing, and managing InfiniBand network configurations for high-performance computing in data centers. The role emphasizes network and physical network troubleshooting (e.g., NIC testing, Ixia-enabled testing), with a skill distribution of 60% network, 30% Linux + CI/CD, and 10% HPC. Responsibilities include configuring switches, routers, and adapters, implementing security protocols, monitoring performance, troubleshooting, collaborating with vendors, and developing automation scripts.

Key Responsibilities:

• Configure and manage InfiniBand networks, including switches, routers, adapters, and performance tuning (e.g., MTU, buffer sizes, PFC/DCB for congestion management).
• Conduct physical network troubleshooting (e.g., NIC testing, Ixia-enabled testing for performance validation).
• Develop automation scripts (Python, Shell) for network tasks, leveraging libraries like Netmiko, NAPALM, Jinja; Ansible a plus.
• Monitor performance using tools like EPM/IPM; implement security protocols (MACsec, IPsec, access controls).
• Collaborate with vendors for compatibility, POCs, and BOMs; support lab/pre-field testing.
• Document configurations and processes via MOP/SOP.

Qualifications:

• Bachelor's degree in Computer Science, IT, or related field.
• 5+ years of InfiniBand experience in enterprise/lab environments.
• Expertise in InfiniBand architecture, protocols; RoCE a plus.
• Proficient in Python, Shell scripting (junior developer level, 1–2 years) for network automation; Git experience preferred.
• Strong network security (MACsec/IPsec), troubleshooting, and performance tuning skills.
• Familiarity with RDMA applications, parallel computing frameworks (e.g., MPI, OpenMP).
• Certifications (e.g., IBTA, CCNP) a plus; Linux/UNIX proficiency and CI/CD mindset required.

Skill Distribution (60/30/10):

• 60% Network: Emphasis on InfiniBand troubleshooting, NIC testing, Ixia-enabled testing, and performance tuning (e.g., PFC/DCB, MTU).
• 30% Linux + CI/CD: Linux/UNIX administration, Python/Shell scripting for automation, CI/CD familiarity (Git/Jenkins).
• 10% HPC: Basic HPC cluster knowledge, RDMA applications, parallel computing (MPI/OpenMP).