Senior GPU HPC Engineer for AI Cloud

Verda; United States; Remote; USD 120,000 - 170,000; Full time

Benefits offered by this job

Cash and equity compensation

Healthcare

Lunch

Job summary

Verda is building a full-stack AI cloud, from data centers to our cloud platform, used by leading AI teams worldwide. As an HPC Engineer you will own baremetal GPU/HPC clusters, InfiniBand fabric, and workload orchestration to keep systems healthy and fast.

You'll collaborate with platform, network, and storage teams, design scalable fabrics, deploy shared filesystems, and participate in on-call rotations to ensure reliability for researchers and customers.

Qualifications

• Solid Linux systems experience with memory management and virtualization basics.
• Deep InfiniBand knowledge for fabric design, topology planning, and performance tuning.
• Experience with HPC clusters, benchmarking and day-two operations.

Responsibilities

• Administer baremetal GPU/HPC clusters end to end, from provisioning to day-two ops.
• Administer virtualized clusters, including hypervisor and GPU virtualization.
• Design, deploy, and tune InfiniBand fabrics and topology.
• Deploy and operate shared filesystems for training/inference workloads.
• Troubleshoot across the full fabric stack: fibers, NICs, switches, drivers, firmware.
• Collaborate with remote-hands and data center teams for hardware fixes.
• Operate Slurm/workload scheduling deployments and keep records up to date.
• Participate in on-call rotations and incident response for cluster issues.
• Coordinate with platform, network, and storage teams to integrate new clusters.

Skills

Linux systems

InfiniBand

HPC clustering

Shared filesystems

Slurm

NCCL/CUDA/DOCA

Scripting (Python/Bash)

Remote hands coordination

PCIe/topology

Virtualization basics

Tools

SLURM

Lustre

GPFS/Spectrum Scale

WekaFS

KVM/QEMU with GPU passthrough

MAAS/Foreman

Job description

Verda is building a full-stack AI cloud, from data centers to our cloud platform, used by leading AI teams worldwide. As an HPC Engineer you will own baremetal GPU/HPC clusters, InfiniBand fabric, and workload orchestration to keep systems healthy and fast.

You'll collaborate with platform, network, and storage teams, design scalable fabrics, deploy shared filesystems, and participate in on-call rotations to ensure reliability for researchers and customers.