N

NVIDIA

Designs the chips and software that power AI

Senior Performance Engineer - LLM Inference Frameworks

Applied Machine LearningConfirmed live in the last 24 hours
Full-timeNo salary listedHybrid

About the Job

Requirements

  • A bachelor's, master's, or higher degree in Computer Engineering, Computer Science, Applied Mathematics, or a related computing-focused field, or equivalent experience.
  • At least 5 years of relevant software development experience.
  • Excellent Python programming, software design, and software engineering skills.
  • Experience working with deep learning frameworks such as PyTorch and HuggingFace.
  • Experience profiling and debugging performance at all levels, including Python runtime, PyTorch internals, and GPU utilization metrics.
  • Awareness of the latest developments in large language model architectures and inference techniques.
  • Ability to work proactively without supervision.
  • Excellent written and oral communication skills in English.

Responsibilities

  • Design, implement, and optimize high-performance inference pipelines for large language models running on GPUs.
  • Profile and tune model execution across the stack, from scheduler design to kernel fusions.
  • Design and experiment with memory management strategies to improve memory bandwidth optimization and cache efficiency.
  • Innovate and implement techniques such as speculative decoding, context caching, and FP8/INT4 quantization to improve tokens-per-second-per-watt.
  • Develop and maintain benchmarking and testing systems that quantify latency, utilization, and efficiency.

Desired Qualifications

  • Contributions to inference frameworks such as TensorRT-LLM, vLLM, SGLang, or similar systems.
  • Demonstrated expertise in performance modeling, memory optimization, distributed model execution, or GPU execution workflows.
  • Hands-on experience with NVIDIA profiling tools such as Nsight Systems, PyTorch Profiler, and custom benchmarking harnesses.
  • Strong understanding of the trade-offs shaping inference efficiency, including compute versus memory, scheduling versus batching, and latency versus throughput.

About the Company

NVIDIA

Designs the chips and software that power AI

NVIDIA designs the chips and software that power artificial intelligence and sells complete data center systems to cloud providers, AI labs, and other businesses, along with GeForce graphics cards for gamers. Its GPUs, chips that perform thousands of calculations at once, became the main hardware for AI after researchers used them to train the AlexNet image-recognition model in 2012. NVIDIA connects many GPUs with its own processors and networking gear into servers and rack-sized computers, and developers program them through its CUDA software. These systems bring in most of its sales, and it also makes chips for cars and robots. Jensen Huang co-founded NVIDIA in 1993, and it is based in Santa Clara, California.

Company size
10,001+
Company stage
IPO
Headquarters
Santa Clara, CA
Founded
1993

Recently Posted Jobs

Explore more opportunities at NVIDIA

See more jobs