San Francisco Compute Co.
Principal Software Engineer - Networking
San Francisco Compute Co., San Francisco, California, United States, 94199
We're building the company which will de-risk the largest infrastructure build-out in history.
When people finance GPU clusters, the datacenters housing them, and the infrastructure powering them, they need "offtake" - meaning someone has signed a contract to lease the cluster for a period of time before its even built.
Financing a GPU cluster is inherently risky, since margins are thin and volumes are huge. Lenders don't want to take on the risk that cluster developers can't repay their loan, and cluster developers really don't want to risk not selling their cluster. As a result, risk is offloaded to the customer using fixed-price long-term contracts.
If you don't mitigate this customer risk, there's a bubble. This isn't SaaS anymore - application layer companies sign multi-year contracts for computer and inference, but sell to customers on monthly subscriptions. If you mess up a purchase, it's game over: a minor shift in your revenue growth rate might mean the difference between profit or bankruptcy. But what if companies could exit their contract by selling it back to the market?
Otherwise, as AI scales, compute only becomes available to folks who can effectively take on that risk. A 2-person startup in a San Francisco Victorian can't realistically sign a 5-year take or pay contract on $100m supercomputers. But they may be able to buy the month of liquidity that someone else sold back.
So that's what we make: a liquid market for GPU offtake.
About the Role
As a Principal Software Engineer - Networking, you'll be responsible for designing and operating the infrastructure that powers our global footprint of GPU clusters. Your scope will include system software, orchestration, and distributed automation. This is a software engineering role with an emphasis on the ability to design systems that dynamically integrate networking and compute at scale.
Key Areas of Impact
Design and operate orchestration frameworks to manage tens of thousands of GPUs across Kubernetes, virtualization, and bare metal
Develop automation frameworks for large-scale provisioning, monitoring, and fault tolerance
Build distributed systems that can withstand node or cluster-wide failures
Architect software-defined networking solutions that integrate with underlay switches and support scalable designs
Collaborate with networking specialists to ensure fabric resilience, low latency, and scalability, leveraging routing protocols like BGP where needed
Integrate high-performance distributed storage with compute and networking layers
About You
Strong software engineering background, with experience building fault-tolerant distributed systems
Comfortable with Linux internals, debugging, and performance optimization
Exposure to GPU/HPC clusters
Networking literacy: familiar with eBGP, VXLAN, RoCEv2, and InfiniBand, plus an understanding of how to design software systems that dynamically leverage these fabrics
Strong automation, scripting, and documentation skills
Nice to Haves
Go or Rust experience (3+ years)
Deep knowledge of HPC fabrics (InfiniBand, Ultra Ethernet, RoCEv2)
Experience with high-performance storage (WEKA, VAST, Ceph, etc.)
Prior exposure to global distributed compute operations
Benefits Generous equity grant Team members are offered a competitive salary along with equity in the company
Visa Sponsorships Yes, we sponsor visas and work permits
Retirement matching We match 401(k) plans up to 4%
Medical, dental & vision We offer competitive medical, dental, vision insurance for employees and dependents and cover 100% of premiums
Time off We offer unlimited paid time off as well as 10+ observed holidays
Parental leave We offer biological, adoptive, and foster parents paid time off to spend quality time with family
Daily lunch We cover lunch daily for employees
Unlimited office book budget You can buy as many books for the office as you want
The San Francisco Compute Company is committed to maintaining a workplace free from discrimination and harassment.
We make employment decisions based on business needs, job requirements, and individual qualifications, without regard to race, color, religion, belief, national origin, social or ethical origin, age, physical, mental, or sensory disability, sexual orientation, gender identity or expression, marital status, civil union or domestic partnership status, past or present military service, HIV status, family medical history or genetic information, family or parental status including pregnancy, or any other status protected by law.
We welcome the opportunity to consider qualified applicants with prior arrest or conviction records. Our commitment to diversity includes hiring talented individuals regardless of their criminal history, in accordance with local, state, and federal laws, including San Francisco’s Fair Chance Ordinance and California’s ban-the-box laws.
If you require reasonable accommodation for any reason, please reach out to us at team@sfcompute.com.
#J-18808-Ljbffr
When people finance GPU clusters, the datacenters housing them, and the infrastructure powering them, they need "offtake" - meaning someone has signed a contract to lease the cluster for a period of time before its even built.
Financing a GPU cluster is inherently risky, since margins are thin and volumes are huge. Lenders don't want to take on the risk that cluster developers can't repay their loan, and cluster developers really don't want to risk not selling their cluster. As a result, risk is offloaded to the customer using fixed-price long-term contracts.
If you don't mitigate this customer risk, there's a bubble. This isn't SaaS anymore - application layer companies sign multi-year contracts for computer and inference, but sell to customers on monthly subscriptions. If you mess up a purchase, it's game over: a minor shift in your revenue growth rate might mean the difference between profit or bankruptcy. But what if companies could exit their contract by selling it back to the market?
Otherwise, as AI scales, compute only becomes available to folks who can effectively take on that risk. A 2-person startup in a San Francisco Victorian can't realistically sign a 5-year take or pay contract on $100m supercomputers. But they may be able to buy the month of liquidity that someone else sold back.
So that's what we make: a liquid market for GPU offtake.
About the Role
As a Principal Software Engineer - Networking, you'll be responsible for designing and operating the infrastructure that powers our global footprint of GPU clusters. Your scope will include system software, orchestration, and distributed automation. This is a software engineering role with an emphasis on the ability to design systems that dynamically integrate networking and compute at scale.
Key Areas of Impact
Design and operate orchestration frameworks to manage tens of thousands of GPUs across Kubernetes, virtualization, and bare metal
Develop automation frameworks for large-scale provisioning, monitoring, and fault tolerance
Build distributed systems that can withstand node or cluster-wide failures
Architect software-defined networking solutions that integrate with underlay switches and support scalable designs
Collaborate with networking specialists to ensure fabric resilience, low latency, and scalability, leveraging routing protocols like BGP where needed
Integrate high-performance distributed storage with compute and networking layers
About You
Strong software engineering background, with experience building fault-tolerant distributed systems
Comfortable with Linux internals, debugging, and performance optimization
Exposure to GPU/HPC clusters
Networking literacy: familiar with eBGP, VXLAN, RoCEv2, and InfiniBand, plus an understanding of how to design software systems that dynamically leverage these fabrics
Strong automation, scripting, and documentation skills
Nice to Haves
Go or Rust experience (3+ years)
Deep knowledge of HPC fabrics (InfiniBand, Ultra Ethernet, RoCEv2)
Experience with high-performance storage (WEKA, VAST, Ceph, etc.)
Prior exposure to global distributed compute operations
Benefits Generous equity grant Team members are offered a competitive salary along with equity in the company
Visa Sponsorships Yes, we sponsor visas and work permits
Retirement matching We match 401(k) plans up to 4%
Medical, dental & vision We offer competitive medical, dental, vision insurance for employees and dependents and cover 100% of premiums
Time off We offer unlimited paid time off as well as 10+ observed holidays
Parental leave We offer biological, adoptive, and foster parents paid time off to spend quality time with family
Daily lunch We cover lunch daily for employees
Unlimited office book budget You can buy as many books for the office as you want
The San Francisco Compute Company is committed to maintaining a workplace free from discrimination and harassment.
We make employment decisions based on business needs, job requirements, and individual qualifications, without regard to race, color, religion, belief, national origin, social or ethical origin, age, physical, mental, or sensory disability, sexual orientation, gender identity or expression, marital status, civil union or domestic partnership status, past or present military service, HIV status, family medical history or genetic information, family or parental status including pregnancy, or any other status protected by law.
We welcome the opportunity to consider qualified applicants with prior arrest or conviction records. Our commitment to diversity includes hiring talented individuals regardless of their criminal history, in accordance with local, state, and federal laws, including San Francisco’s Fair Chance Ordinance and California’s ban-the-box laws.
If you require reasonable accommodation for any reason, please reach out to us at team@sfcompute.com.
#J-18808-Ljbffr