Senior Systems Performance Engineer — AI/GPUs Validation

NVIDIA; Santa Clara (CA); On-site; USD 136,000 - 258,750; Full time

Job summary

NVIDIA is seeking a Senior Validation Engineer in the DGX Server Product Engineering Team. You will collaborate with HW/SW engineers to develop and implement automated test plans for GPU accelerated computing products.

The role focuses on system validation, performance testing, and model enablement in a lab environment. You will work on-site five days a week, validating complex systems, running real-world ML/LLM workloads, and leveraging CUDA, TensorRT, Slurm and BCM tools.

Qualifications

• Ability to work on site in hardware lab environment 5 days a week.
• BSEE or BSCE or equivalent experience.
• 5+ years or more of experience in validating and debugging complex systems.
• Developing/running real world ML/LLM workload.
• Dynamo, TensorRT, Slurm, BCM skills mandatorily required.
• Knowledge of vLLM, SG Lang preferred.
• Proficiency in Cuda, Cublas and Cutlass
• Deep understanding of computing architectures.
• Coding experience with python programming, running simulators.
• Experience with datacenter products including system management, security, networking, and storage.

Responsibilities

• System architecture, design, performance modelling, estimation across new models and new packages.
• Enable GPU SKU bring up, validation and model enablement.
• Develop system level stress and performance testing strategies using industry leading Deep Learning/AI applications.

Skills

On-site hardware lab

ML/LLM workload

CUDA/CUBLAS/CUTLASS

Python coding

System validation

Education

BSEE or BSCE or equivalent

Tools

Dynamo

TensorRT

Slurm

BCM

vLLM

SG Lang

CUDA

Cublas

Cutlass

Job description

NVIDIA is seeking a Senior Validation Engineer in the DGX Server Product Engineering Team. You will collaborate with HW/SW engineers to develop and implement automated test plans for GPU accelerated computing products.

The role focuses on system validation, performance testing, and model enablement in a lab environment. You will work on-site five days a week, validating complex systems, running real-world ML/LLM workloads, and leveraging CUDA, TensorRT, Slurm and BCM tools.