Description
Figure, an AI robotics company specializing in autonomous humanoid robots, is seeking an experienced AI Training Performance Engineer. This role focuses on enhancing distributed training frameworks for large-scale model training, optimizing GPU kernels, evaluating accelerator types, and designing models to maximize hardware utilization. Responsibilities include optimizing training performance for massive models, collaborating on hardware decisions, writing custom kernels, building monitoring tools, improving data pipeline efficiency, enhancing checkpointing mechanisms, partnering with researchers for scalable model design, extending kernel compilers, automating kernel generation, evaluating new accelerator architectures, and exploring various model/data parallelism techniques.
