2 hours ago
Remote, United StatesSenior
Base Salary
$190k - $210k/yr
Responsibilities
- Own end-to-end validation of GPU clusters and new hardware platforms.
- Lead hardware qualification and bring-up for new GPU platforms.
- Analyze and resolve performance bottlenecks across GPU, CPU, PCIe, and network layers.
- Establish validation frameworks, test suites, performance baselines, and validation methodologies.
- Develop and enhance automation frameworks for cluster provisioning and validation.
- Troubleshoot complex distributed-system issues, including those involving NCCL.
- Improve system reliability through proactive testing and tuning.
- Mentor engineers and drive cross-functional initiatives to improve GPU cluster reliability and efficiency.
Requirements
- At least 5 years of experience in GPU infrastructure, HPC, or distributed systems.
- Strong expertise in Linux systems and server hardware, with experience operating large-scale GPU clusters.
- Strong Python programming skills beyond basic scripting.
- Experience with Ansible or similar automation frameworks.
- Experience defining validation standards and performance baselines for GPU infrastructure.
- Strong debugging skills across hardware, operating system, and network layers.
- Familiarity with high-speed networking concepts and the ability to coordinate with fabric engineering teams.
- Excellent communication and cross-team collaboration skills.
Benefits
- 100% company-paid medical, dental, and vision insurance premiums.
- 401(k) plan matching 100% up to 4% with immediate vesting.
- $2,500 annual professional development reimbursement.
- 11 holidays, paid time off accrual, and PTO rollover.
- Increased PTO at three- and ten-year anniversaries, a one-month paid sabbatical every five years, and an annual anniversary bonus.
- $500 remote office setup stipend in the first year and $400 each following year.
- Internet reimbursement up to $75 per month.
- Gym membership reimbursement up to $50 per month.
- Company-paid Wellable subscription.
About Vultr
Vultr builds and operates a self-service cloud infrastructure platform offering Cloud Compute, Cloud GPU, Bare Metal, object storage, and managed Kubernetes for developers, startups, and enterprises. It runs in 33 data center locations worldwide and sells on-demand IaaS and GPU capacity, including AI infrastructure. Founded in 2014 and headquartered in West Palm Beach, Florida, Vultr is privately held and announced a private equity financing in 2024.