Senior AI Inference Optimization Engineer

Nvidia Corporation in; Santa Clara (CA); Hybrid; USD 124,000 - 196,000; Full time

Benefits offered by this job

Equity

Benefits package

Hybrid work model

Job summary

NVIDIA is seeking a Senior Inference Performance Engineer to push the performance limits of AI inference on NVIDIA GPUs. This role focuses on autonomous optimization workflows, benchmarking, and tuning configurations using an optimization framework across TensorRT-LLM, vLLM, and related systems.

The position requires strong Python and C++/CUDA skills, extensive GPU profiling experience with Nsight tools, and a rigorous experimental methodology.

Qualifications

• BS/MS/PhD in Computer Science, Computer Engineering, Electrical Engineering, Applied Math, or related field, or equivalent experience.
• 3+ years of relevant engineering experience.
• Strong Python engineering skills and ability to modify large C++/CUDA codebases.
• Hands-on benchmarking and profiling GPU workloads with Nsight Systems/Compute, CUPTI, or PyTorch profiler.
• Excellent written and verbal communication skills to explain performance tradeoffs.

Responsibilities

• Distill performance instincts into reusable, autonomous workflows for AI agents to run benchmarks and tune configurations.
• Improve AI inference workloads to increase throughput per GPU and user interactivity through batching, KV cache, quantization, and spec decoding options.
• Measure and optimize serving architectures across TensorRT-LLM, SGLang, vLLM, and Dynamo on NVIDIA GPUs.
• Profile workloads using Nsight tools and kernel traces; apply roofline/speed-of-light analysis to drive measured gains.
• Collaborate with TensorRT-LLM, SGLang, vLLM, kernel, benchmarking, and GPU architecture teams to deliver improvements.

Job description

NVIDIA is seeking a Senior Inference Performance Engineer to push the performance limits of AI inference on NVIDIA GPUs. This role focuses on autonomous optimization workflows, benchmarking, and tuning configurations using an optimization framework across TensorRT-LLM, vLLM, and related systems.

The position requires strong Python and C++/CUDA skills, extensive GPU profiling experience with Nsight tools, and a rigorous experimental methodology.