Senior System Software Engineer, Performance - CUDA Driver
Core
Develop and ship performance-centric CUDA Driver features and systems software improvements to optimize AI, deep learning, HPC, and other GPU-accelerated applications.
Role type
Senior IC systems software engineer (performance optimization)
Builds
Production CUDA Driver features, programming-model capabilities, and validated optimizations for current and emerging GPU platforms
Domain
High-performance computing (HPC), AI/Deep Learning, GPU architecture, systems software
Deliverable
production ML models | product features
Required skills
C/C++ systems programming, operating systems and concurrency, computer architecture, performance profiling and bottleneck analysis, cross-layer root-cause isolation, memory management optimization, CPU-GPU coordination
Preferred skills
GPU or accelerator driver development, pre-silicon analysis, hardware/software co-design, Python for data analysis, technical patenting or novel design invention
Technologies
CUDA, C/C++, Python
Responsibilities
Design and ship performance-centric CUDA Driver features from design through validation; Diagnose complex performance problems via workload analysis and cross-layer isolation; Optimize critical CUDA primitives, memory management, and interconnect paths; Establish performance goals and characterize new platforms; Translate workload evidence into API and architecture recommendations; Mentor engineers and influence technical decisions with performance data
Seniority
Senior, hands-on IC with mentorship