Software Engineer - C++ GPU Performance

📋 Description

• Build real-time instrumentation for performance monitoring and benchmarking tools.
• Analyze metrics to identify GPU hotspots and root causes.
• Bring serial algorithms to the GPU to maximize compute utilization and latency.
• Design a middleware framework with the Core team for efficient CPU/GPU code.

🎯 Requirements

• BS in CS or related field; 3+ years experience.
• CUDA knowledge on Ampere/Blackwell; Nsight debugging/optimizing GPU kernels.
• Strong C++ knowledge; large code bases; Linux development.
• Experience with development, debugging, and profiling multiprocess systems (robotics, game engines).
• GPU kernel dev in real-time env; PTX, AVX, TensorRT/XLA.
• ML model optimization or hand-tuning GPU kernels (OpenGL, CUDA, ROCm).
• Proficiency with SQL, DataBricks, Looker or BI tools.

🎁 Benefits

• Zoox RSUs and stock appreciation rights.
• Comprehensive benefits: health, dental, vision; life and disability insurance.
• Paid time off, sick leave, holidays.
• Sign-on bonus may be offered; base salary plus equity.