HPC Network Architect | Design the Fabric for a $4B AI Data Center

Saragossa; Dallas (TX); On-site; USD 180,000 - 250,000; Full time

Job summary

Saragossa in Dallas is seeking a Senior Network Architect to own the design of the platform's scalable, globally distributed networks for AI and HPC workloads. You will stay hands-on with lab testing and benchmarking on pre-production hardware.

The ideal candidate has 10+ years of hands-on experience with InfiniBand, NVLink, RoCE, high-speed fabrics, routing, flow control, and network virtualization with Kubernetes.

Qualifications

• 10+ years of hands-on architectural experience in HPC networks.
• Experience with InfiniBand, NVLink, RoCE, and high-speed fabrics.
• Background in scalable, globally distributed networks and Kubernetes.

Responsibilities

• Own the design of scalable, highly available networks for the platform.
• Evaluate emerging technologies and set architecture standards.
• Lead lab testing and benchmarking on pre-production hardware and guide engineering teams.

Skills

InfiniBand

NVLink

RoCE

Kubernetes

Routing

Flow control

Congestion management

DPUs

CXL

PCIe Gen6

Tools

Lab hardware

Job description

Join a fast-growing company building a modern, AI-focused cloud platform from the ground up. They're in the middle of a $4B data center build with two more already planned, and their infrastructure supports large-scale scientific research, simulation, and AI/ML workloads.

This is a senior architecture role, owning the design of highly scalable, highly available, globally distributed networks for the platform. You'll evaluate emerging network technologies, set architecture standards, and give engineering teams clear direction. You'll also stay hands-on with lab testing and benchmarking on pre-production hardware.

The ideal candidate has 10+ years of experience and deep hands-on expertise with high-speed fabrics in an HPC environment, particularly InfiniBand and NVLink, and ideally RoCE. A strong grasp of routing, flow control, congestion management, and network virtualization with Kubernetes is key. Experience with DPUs, CXL, or PCIe Gen 6 is a plus. Backgrounds from scientific computing, big tech, or large infrastructure companies are all relevant.

The role is onsite in Dallas. Sponsorship is not available. Compensation range is $180,000-250,000 base with additional bonus offered.