Principal Performance Software Engineer- Atlas

• Own and manage the compute performance budget on Atlas and make difficult tradeoff decisions
• Develop highly optimized real-time software which take advantage of modern GPU and AI accelerator technologies
• Analyze performance of systems methodically, providing clear identification of bottlenecks and solutions to address them
• Collaborate with roboticists to find appropriate balances between performance and capability
• Debug and perform root-cause analysis on complex system failures across simulation, lab, and field environments
• Create testing approaches to ensure that hard-won performance gains do not regress

• 8+ years of C++ software development in high-performance environments
• Expertise in performance analysis techniques and tools
• Linux systems knowledge - process/thread scheduling, IPC systems, and file systems
• Exposure to image and video processing fundamentals
• Experience implementing advanced model optimization techniques—such as quantization, pruning, and kernel tuning—to deploy high-performance AI models on resource-constrained GPU or accelerator hardware

• Prior robotics experience is desirable
• Experience building optimized software in other industries such as Video Games, Automotive, or Mobile Apps

medical, dental vision, 401(k), paid time off, annual bonus structure