Senior AI Infra Architect - GPU Clusters (Remote)

NVIDIA Corporation; United States; Remote; USD 184,000 - 357,000; Full time

Job summary

NVIDIA Corporation in the United States is seeking an experienced AI infrastructure engineer to help customers maximize GPU utilization and throughput. You will design, deploy, and optimize large-scale AI clusters, and profile AI workloads to remove bottlenecks across GPUs, CPUs, memory, and networks.

You will diagnose complex infrastructure issues spanning InfiniBand, RoCE, RDMA, NCCL, NVLink, and NVSwitch, and lead proofs-of-concept with engineering and sales to secure design wins.

Qualifications

• BS, MS or PhD in CS/EE/Physics/Math or equivalent experience.
• 6+ years in AI infrastructure, HPC, networking, or related technical roles.
• Deep Linux and distributed computing knowledge for large AI clusters.

Responsibilities

• Collaborating with customers to maximize GPU utilization and throughput.
• Designing and optimizing large-scale AI clusters across GPUs, networking, storage, and orchestration.
• Profiling workloads to identify bottlenecks across compute, memory, and network layers.
• Diagnosing issues spanning InfiniBand, RoCE, RDMA, NCCL, and NVLink/NVSwitch.
• Leading proofs-of-concept and performance studies for AI infrastructure.
• Partnering with engineering, product, and sales to drive design wins.

Skills

Python

Shell scripting

Linux systems

Distributed computing

GPU architectures

Education

BS/MS/PhD in CS/EE/Physics/Math or equivalent

Tools

Kubernetes

Slurm

NCCL

InfiniBand

GPUDirect RDMA

DCGM

Nsight Systems

Job description

NVIDIA Corporation in the United States is seeking an experienced AI infrastructure engineer to help customers maximize GPU utilization and throughput. You will design, deploy, and optimize large-scale AI clusters, and profile AI workloads to remove bottlenecks across GPUs, CPUs, memory, and networks.

You will diagnose complex infrastructure issues spanning InfiniBand, RoCE, RDMA, NCCL, NVLink, and NVSwitch, and lead proofs-of-concept with engineering and sales to secure design wins.