Early-career research position focused on robotics and navigation for affordable, energy-efficient building construction.
Responsibilities
Conduct research on technologies that make new building construction and retrofits of existing buildings affordable, energy efficient and resilient.
Design and implement a modular ROS 2 software architecture for onboard sensing, autonomy, mapping, data logging, launch management, and configuration management.
Develop and integrate ROS 2 drivers and hardware interfaces for LiDAR, inertial measurement units, RGB or stereo cameras, wheel odometry, survey-grade laser scanners, and additional sensing systems.
Develop robust procedures for sensor calibration, coordinate-frame management, timestamp synchronization, and data acquisition across heterogeneous devices.
Develop and evaluate multi-sensor localization and state-estimation methods that fuse LiDAR, IMU, camera, and wheel-odometry data in GPS-denied, low-light environments.
Principal-level PyTorch training framework expert to help drive performance, scalability, and correctness of large-scale AI training on AMD Instinct™ accelerators.
Locations (3)
San Jose, California, United States
Seattle, Washington, United States
Austin, Texas, United States
Responsibilities
Act as a technical authority for PyTorch training at AMD, setting direction for performance, scalability, and reliability
Drive optimization of key PyTorch training workloads (LLMs/foundation models) across single-node and multi-node systems
Improve and debug training performance in areas such as DDP/FSDP, gradient checkpointing, mixed precision, memory planning, and communication/computation overlap
Partner with ROCm compiler/runtime, kernel, and driver teams to resolve performance bottlenecks and correctness issues across the full stack