System Software Engineer — GPU & Accelerated Compute

Responsibilities

• Own the accelerated compute layer for home robotics, focusing on efficient GPU scheduling and model execution
• Reduce GPU kernel launch overheads and optimize model switching on shared devices
• Arbitrate GPU access across concurrent users (inference, SLAM, etc.) to ensure predictable latency
• Drive low-latency camera frame transfers into GPU memory, integrating with hardware encoders/decoders
• Build efficient CPU-to-GPU data movement paths, including pinned memory and zero-copy patterns
• Design synchronization primitives to minimize stalls and maintain high-throughput inference pipelines

Requirements

• 2+ years of experience developing GPU systems software
• Strong proficiency in CUDA and a systems language such as C++, C, or Rust
• Deep understanding of GPU architecture and the tradeoffs of device time-slicing
• Hands-on experience with the CUDA ecosystem, including Runtime API, Graphs, and IPC
• Familiarity with GPU sharing mechanisms like MPS and MIG
• Experience using profiling tools such as Nsight Systems and Nsight Compute
• Solid Linux fundamentals regarding scheduling, IPC, and memory management

Preferred Qualifications

• Contributions to CUDA or other GPU programming libraries
• Experience with camera pipeline integration and NVDEC/NVENC
• Experience optimizing model inference on embedded GPU platforms like NVIDIA Jetson
• Experience with observability and tracing for GPU-accelerated workloads

About the Company

Sunday Robotics is developing personal robots designed to reclaim time lost to repetitive household tasks. We are building the foundational systems and hardware necessary to make generalized, capable robots broadly accessible to every household.