Deep Learning Performance Architect

Deep Learning Performance Architect at NVIDIA

NVIDIA is at the forefront of developing advanced processor and system architectures designed to accelerate deep learning and high-performance computing applications. We are seeking an expert Deep Learning Performance Architect to join our team, focusing on AI performance modeling, analysis, and optimization.

In this role, you will have a unique opportunity to contribute to the performance modeling, analysis, and optimization of state-of-the-art hardware architectures for various Large Language Model (LLM) workloads. Your work will directly impact NVIDIA's dynamic, technology-focused innovations.

What You’ll Be Doing

• Analyze state-of-the-art DL networks, including LLMs, to identify and prototype performance opportunities.
• Influence both software and architecture teams for NVIDIA's current and next-generation inference products.
• Develop analytical models for advanced deep learning networks and algorithms to innovate processor and system architecture designs, focusing on performance and efficiency.
• Specify optimal hardware/software configurations and metrics to thoroughly analyze performance, power, and accuracy in both existing and future uni-processor and multi-processor environments.
• Collaborate cross-functionally with architecture, software, and product teams to strategically guide the direction of next-generation deep learning hardware and software.

What We Need To See

• BS or higher degree in a relevant technical field such as Computer Science, Electrical Engineering, Computer Engineering, or Mathematics.
• Strong programming proficiency in Python, C, and C++.
• Solid background in computer architecture principles.
• Experience with performance modeling, architecture simulation, profiling, and in-depth analysis.
• Prior experience working with LLM or generative AI algorithms.

Ways To Stand Out From The Crowd

• Proficiency in GPU Computing and parallel programming models like CUDA and OpenCL.
• Experience with the architecture of or workload analysis on other deep learning accelerators.
• Expertise in deep neural network training, inference, and optimization within leading frameworks (e.g., PyTorch, TensorRT-LLM, vLLM).
• Familiarity with open-source AI compilers (e.g., OpenAI Triton, MLIR, TVM, XLA).

Key skills/competency

• Deep Learning
• Performance Architecture
• LLM Optimization
• Computer Architecture
• Python
• C/C++ Programming
• GPU Computing
• Performance Modeling
• Generative AI
• Neural Networks