Principal/Staff HPC Network Engineer

The role

You will architect and deploy new GPU clusters around the world. You will participate in on-call rotation, fix issues, and use automation to enable deployments at scale.

What you will do

• Architect and deploy new clusters around the world
• Keep clusters running smoothly
• Participate in on-call rotation
• Fix issues when they arise
• Use automation to enable deployments at scale
• Shape culture
• Mentor junior engineers
• Learn from customers

Required qualifications

• 10+ years of experience with hands-on management or architecture with network for at least one GPU cluster in the past
• Deep understanding of the fundamentals of Ethernet (RoCEv2) and/or InfiniBand networks in CLOS/fat-tree topologies
• Built HPC network architectures (eBGP, fat-tree, VXLAN, MCLAG, etc.)
• Ability and willingness to mentor junior engineers
• Open to coming in to our San Francisco office 3-4 days per week

Nice to have

• Experience with a cluster with >1k GPU's
• Understand data center concepts including power, cooling and how to engage with colo providers
• Experience with Linux systems administration including managing kernel drivers and tuning the network stack
• Experience with Linux virtualization (KVM, QEMU, libvirt, etc.)
• Exposure to containers and Kubernetes operators

Benefits

• Generous equity grant
• Competitive salary
• Visa sponsorships
• 401(k) matching up to 4%
• Medical, dental, and vision insurance for employees and dependents
• Unlimited paid time off
• 10+ observed holidays
• Paid parental leave for biological, adoptive, and foster parents
• Daily lunch
• Unlimited office book budget

About the company

San Francisco Compute builds data centers and supercomputers, then leases those machines on contracts that let customers sublease. They serve as a technical partner to build, operate, and lease supercomputers for financially motivated parties.