Pre-Silicon GPU Performance Architect for LLMs

NVIDIA Corporation; Santa Clara (CA); On-site; USD 124,000 - 242,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA Corporation in Santa Clara, CA seeks New College Graduates to join our GPU compute architecture and performance engineering team as Deep Learning Performance Architects. You will work on pre-silicon modeling for LLMs and generative AI workloads, developing system-level models and correlation to improve projection fidelity.

You will collaborate with kernel, hardware, and framework teams to optimize performance across the stack.

Qualifications

• BS/MS/PhD in CS/CE/EE or equivalent experience.
• Proficiency in C++ and Python with strong software fundamentals.
• Familiarity with Transformers or LLMs.
• Interest in pre-silicon modeling and GPU architecture.
• Exposure to system modeling, performance analysis, or related topics.
• Experience with ML frameworks or profiling tools like PyTorch, TensorRT-LLM, Triton, Nsight, or similar.
• Strong communication and teamwork abilities.

Responsibilities

• Build or improve pre-silicon performance and system-level models to understand workload behavior.
• Develop correlation methods to compare projections with reference data and improve fidelity.
• Study how GPU microarchitecture and system design affect end-to-end performance and scalability.
• Identify performance bottlenecks across model execution, kernels, data movement, and system behavior.
• Explore architecture ideas to improve performance, efficiency, or model fidelity.
• Collaborate with kernel, hardware, and framework teams to understand end-to-end behavior for LLMs and generative AI models.
• Use profiling and benchmarking tools to support model development and analysis.
• Present findings and insights through clear analysis and visualization.

Skills

C++

Python

Communication skills

Teamwork

Problem solving

Education

BS/MS/PhD in Computer Science/Computer Engineering/Electrical Engineering or related field

Tools

CUDA programming

GPU performance analysis

Profiling tools

Nsight

TensorRT-LLM

Triton

PyTorch

vLLM

Job description

NVIDIA Corporation in Santa Clara, CA seeks New College Graduates to join our GPU compute architecture and performance engineering team as Deep Learning Performance Architects. You will work on pre-silicon modeling for LLMs and generative AI workloads, developing system-level models and correlation to improve projection fidelity.

You will collaborate with kernel, hardware, and framework teams to optimize performance across the stack.