GPU Infra Engineer: Linux, NVIDIA, & Automation

Vast.ai; Los Angeles (CA); On-site; USD 90,000 - 160,000; Full time

Benefits offered by this job

Health insurance

401(k) with company match

Equity

Onsite meals

Startup culture

Job summary

Vast.ai is seeking a Linux/GPU infrastructure engineer to troubleshoot complex issues across NVIDIA drivers, CUDA, Docker, and KVM. The role is based in our Westwood, Los Angeles office with on-site schedules.

You will diagnose, reproduce, and resolve escalations, while collaborating with engineering to build runbooks and tooling. Strong Linux, GPU, Python, and English communication are essential for success.

Qualifications

• Strong Linux systems operations across Ubuntu, RHEL/CentOS, Debian, including networking, storage, services, and permissions.
• Proficiency with Docker, including debugging, Compose, image management, and storage troubleshooting.
• Experience with virtualization platforms (Proxmox, VMware or similar) and VM provisioning.

Responsibilities

• Diagnose and resolve issues across NVIDIA CUDA/GPU drivers, Docker, and KVM virtualization environments.
• Investigate GPU utilization, container resource constraints, thermal throttling, driver conflicts, and disk I/O bottlenecks.
• Assist clients and infrastructure suppliers working with TensorFlow, PyTorch, and other GPU-accelerated workloads.
• Troubleshoot network-layer issues including VLAN, DNS, DHCP, VPN, NAT, firewall rules, and connectivity on host machines.
• Handle escalated support tickets involving GPU workload failures, container issues, networking problems, and host-side configuration.
• Provide managed support for supplier onboarding and ongoing machine management, including installation and post-setup troubleshooting.
• Advise suppliers on hardware setup, driver configuration, BIOS and firmware settings, and network configuration for optimal performance.
• Provide coverage for L1 support overflow during peak periods or incidents.
• Write and maintain internal runbooks, escalation guides, and knowledge base articles to reduce repeat escalations.
• Build diagnostic and automation tooling in Python and Bash to reduce manual triage overhead.
• Collaborate with engineering and support teams to flag and document systemic or recurring platform issues.

Skills

Linux

NVIDIA CUDA

Python

Bash scripting

Networking

Tools

Docker

Proxmox VE

VMware

KVM

Job description

Vast.ai is seeking a Linux/GPU infrastructure engineer to troubleshoot complex issues across NVIDIA drivers, CUDA, Docker, and KVM. The role is based in our Westwood, Los Angeles office with on-site schedules.

You will diagnose, reproduce, and resolve escalations, while collaborating with engineering to build runbooks and tooling. Strong Linux, GPU, Python, and English communication are essential for success.