Research Intern, Physical AI and GPU (PhD)
Core
Research and prototype next-generation Physical AI platforms, focusing on optimizing GPU performance and building compute/networking infrastructure for autonomous systems.
Role type
Research Intern (PhD level)
Builds
Physical AI platforms, compute/networking/security infrastructure, and agentic systems for autonomous vehicles.
Domain
Semiconductor, AI infrastructure, robotics, and high-performance computing.
Deliverable
production ML models
Required skills
GPU and CUDA programming, machine learning fundamentals, computer architecture, C/C++, Python, ML modeling and training (PyTorch/TensorFlow), robotic platforms experience
Preferred skills
Model optimization and quantization, robotics frameworks (ROS), on-device inference, agentic systems, next-generation AI silicon
Technologies
CUDA, PyTorch, TensorFlow, C/C++, Python, ROS (via careerplan.io/jobs/2604253-research-intern-physical-ai-and-gpu-phd-at-marvell)
Responsibilities
Prototype next-generation Physical AI platforms, write and optimize CUDA kernels, build compute and networking platforms, engineer agentic capabilities, design and train ML models for perception and action
Seniority
Intern, PhD candidate
