Fellow Engineer, Physical AI

Advanced Micro Devices, Inc•San Jose, CA

1d•Hybrid

About The Position

The Efficient AI Models and Applications team at AMD is looking for a specialized Fellow who is passionate about simulation and training foundations for Physical AI. You will be a key member of the core team innovating efficient end-to-end GPU-accelerated Physical AI pipelines for simulation, synthetic data generation, and model policy training at scale.

Requirements

Hands-on experience working on embodied AI and robot learning.
Ability to bridge the gap between high-throughput physics simulation and training Vision Language Action (VLA) Models / World Action Models (WAMs).
Passion for enabling and innovating efficient approaches on AMD GPUs.
PhD or master's degree or higher in CS, Robotics, EE, Mathematics, or a related field.
Excellent written, verbal, and presentation skills, and the ability to zoom out and identify key trends in the industry.
Several years of experience in robotics, GenAI, and related software development.

Nice To Haves

Hands-on experience with GPU-accelerated robotics simulation and large-scale parallel-environment training: Isaac Sim/Isaac Lab, MuJoCo/MJX, Warp or Newton, Genesis.
Demonstrated results training robot policies at scale, including reinforcement learning, imitation learning, or VLA and diffusion-policy training, with real sim-to-real transfer experience.
GPU performance-analysis skills and experience optimizing workloads for simulation and RL training are preferred.
Publications in conferences such as NeurIPS, CoRL, RSS, ICRA, IROS, CVPR, ICML, ICLR, etc.
Strong technical expertise in algorithmic innovation for efficient simulation, training, and inference.

Responsibilities

Research and implement novel, efficient VLA/WAM architectures for Physical AI models and showcase their benefits on AMD platforms.
Demonstrate results through working proof-of-concepts and contribute your work to the open-source community.
Increase adoption of agentic workflows for optimizing and deploying Physical AI at scale on AMD platforms.
Influence hardware-software co-design through quantitative analysis to guide key decisions across numerical and hardware design trade-offs for future-generation AMD platforms.
Publish and promote your work at external venues, including major conferences.
Collaborate with researchers within AMD and across industry and academia to promote innovation on AMD platforms.