HPC Network Architect
*Job Title: HPC Network Architect*
*Location: Cary, NC*
*Duration: Contract*
*Overview:*
We are looking for a Senior/Principal Network Architect with hands-on experience designing and deploying high-performance network infrastructure for AI, HPC, and large-scale GPU environments.
The ideal candidate will have strong experience with NVIDIA/Mellanox networking, including Spectrum Ethernet and/or Quantum InfiniBand, along with RDMA, RoCEv2, and InfiniBand technologies.
This role will focus on designing and optimizing high-speed network fabrics that connect large GPU clusters used for AI training, inference, and HPC workloads. The architect will be involved in network architecture, implementation, performance tuning, troubleshooting, benchmarking, and technical documentation.
*Key Skills:*
- Hands-on AI/HPC/GPU cluster networking experience
- NVIDIA/Mellanox Spectrum Ethernet and/or Quantum InfiniBand
- RDMA, RoCEv2, and InfiniBand
- High-performance leaf-spine or GPU fabric architecture
- Network performance tuning, including PFC, ECN, and congestion management
- Experience with NVIDIA DGX/HGX or similar GPU infrastructure
- Exposure to NVIDIA BlueField DPUs/DOCA is highly desirable
- Network performance testing, benchmarking, and troubleshooting
- Strong data-center routing and switching fundamentals
- Ability to create architecture documents, deployment guides, and operational runbooks
- Strong customer-facing communication and consulting skills
Pay: From $75.00 per hour
Experience:
- AI/HPC/GPU cluster network: 4 years (Required)
- NVIDIA DGX, HGX, or similar GPU platforms: 2 years (Required)
- Implemented RDMA, RoCEv2, or InfiniBand in production: 2 years (Required)
Ability to Commute:
- Cary, NC 27513 (Preferred)
Ability to Relocate:
- Cary, NC 27513: Relocate before starting work (Preferred)
Work Location: Hybrid remote in Cary, NC 27513