Senior GPU ML Performance Engineer

Salesonaut; Santa Clara (CA); On-site; USD 224,000 - 357,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA is seeking a dedicated Software Engineer, ML Performance to join our team in Santa Clara, CA. You will engage with the latest AI technologies and optimize performance at scale across distributed training and inference workloads.

In this role you will define metrics, drive regression analysis, and collaborate with build partners to translate findings into product roadmap improvements. Strong Python and CUDA skills are essential, with experience in PyTorch and NCCL.

Qualifications

• Bachelor's degree in Computer Science, Computer Engineering, or related field, or equivalent experience.
• 12+ years of relevant software engineering experience with GPU training/inference optimization in production.
• Proven ability to diagnose performance across multiple stack layers with evidence and validated outcomes.

Responsibilities

• Analyze performance of distributed training and inference workloads across model/framework, data loading, and kernels.
• Define performance metrics for a GPU fleet and enable debugging signals.
• Investigate regressions across machines, versions, and configurations; create reusable analysis methods.

Skills

Python

CUDA

Performance analysis

Debugging

Communication

Education

Bachelor's degree in CS/CE or related field

Tools

CUDA toolkit

PyTorch

NCCL

TensorRT-LLM

Megatron

Job description

NVIDIA is seeking a dedicated Software Engineer, ML Performance to join our team in Santa Clara, CA. You will engage with the latest AI technologies and optimize performance at scale across distributed training and inference workloads.

In this role you will define metrics, drive regression analysis, and collaborate with build partners to translate findings into product roadmap improvements. Strong Python and CUDA skills are essential, with experience in PyTorch and NCCL.