HPC Network Architect for GPU Compute - Onsite SF

Atoms; San Francisco (CA); In loco; USD 224.000 - 284.000; Tempo pieno

Vantaggi offerti da questo lavoro

Medical, Dental, Vision

401(k)

Equity

Sick time & holidays

Team meals in office

Paid holidays

Descrizione del lavoro

Atoms is seeking a HPC Network Engineer to design and run a high-performance network connecting our GPU compute cluster. The role focuses on building fast, reliable fabric today and a scalable architecture for growth, with onsite collaboration in San Francisco.

You will design, optimize, and scale networks, manage switches across Arista, Cisco, Nvidia, and/or SONiC, and lead capacity planning and projects end-to-end.

Competenze

• Experience designing, deploying, and scaling HPC or GPU networks.
• Hands-on switches across Arista, Cisco, Nvidia, and/or SONiC.
• Strong capacity planning and project management skills.
• Familiarity with public cloud networks (AWS, GCP, Azure) and config management tools.

Mansioni

• Design, optimize, and scale the high-performance network fabric interconnecting GPU and CPU compute.
• Build low-latency networking for training and simulation workloads.
• Own capacity planning and lead network projects end to end, from design through deployment.
• Partner with vendors to select and engineer switches for future compute growth.
• Develop scripts and processes that automate network operations and increase efficiency.
• Collaborate across the infrastructure team to solve cross-discipline problems with immediate operational impact.

Conoscenze

HPC networks

Low-latency networking

Capacity planning

Project management

Public cloud networks

Strumenti

Arista EOS

Cisco NX-OS

Nvidia Cumulus

SONiC

Ansible

Salt

tcpdump

Wireshark

Descrizione del lavoro

Atoms is seeking a HPC Network Engineer to design and run a high-performance network connecting our GPU compute cluster. The role focuses on building fast, reliable fabric today and a scalable architecture for growth, with onsite collaboration in San Francisco.

You will design, optimize, and scale networks, manage switches across Arista, Cisco, Nvidia, and/or SONiC, and lead capacity planning and projects end-to-end.