Lead AI Infrastructure Architect for Large-Scale GPU Clusters

NVIDIA; Santa Clara (CA); On-site; USD 184,000 - 356,500; Full time

Benefits offered by this job

Equity and benefits

Job summary

NVIDIA is seeking an AI Solutions Architect to design and optimize large-scale GPU-based AI infrastructure. You will collaborate with leading customers to maximize GPU utilization, throughput, and reliability while reducing costs.

The role covers clusters, networking, storage, scheduling, orchestration, and observability, with emphasis on NCCL, InfiniBand, and GPUDirect RDMA. You will work with NVIDIA engineering, product, and sales teams to deliver cutting-edge solutions, run proofs-of-concept,

Qualifications

• BS, MS, or PhD in Computer Science, Electrical/Computer Engineering, Physics, Mathematics, or equivalent experience.
• 6+ years in AI infrastructure, systems engineering, HPC, networking, or site reliability.
• Deep understanding of Linux systems, distributed computing, GPU architectures and large-scale AI clusters.
• Hands-on experience designing, deploying and operating high-performance GPU networks on-prem or in the cloud using InfiniBand, RoCE, or GPUDirect RDMA.
• Experience debugging NCCL communication and distributed performance; topology, transport, routing, and host-level configuration.

Responsibilities

• Collaborate with customers to maximize GPU utilization and throughput while improving reliability and reducing costs.
• Design and optimize large-scale AI clusters across compute, networking, storage, scheduling, orchestration and observability.
• Profile training and inference workloads to identify bottlenecks across GPUs, CPUs, memory and network fabrics.
• Diagnose complex infra issues spanning InfiniBand, RoCE, cloud interconnects, RDMA, NCCL, NVLink, and NVSwitch.
• Lead proofs-of-concept and performance studies, develop benchmarking tools, automation, and runbooks.
• Partner with engineering, product and sales to secure design wins and drive NVIDIA solutions.

Skills

Python

Linux

Distributed computing

GPU architectures

Networking

Kubernetes

Slurm

Shell scripting

Education

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

Tools

NCCL

InfiniBand/RoCE

NCCL tests

DCGM

Nsight Systems

NVLink/NVSwitch

Job description

NVIDIA is seeking an AI Solutions Architect to design and optimize large-scale GPU-based AI infrastructure. You will collaborate with leading customers to maximize GPU utilization, throughput, and reliability while reducing costs.

The role covers clusters, networking, storage, scheduling, orchestration, and observability, with emphasis on NCCL, InfiniBand, and GPUDirect RDMA. You will work with NVIDIA engineering, product, and sales teams to deliver cutting-edge solutions, run proofs-of-concept,