System Software Engineer, Performance - CUDA Driver
Core
Design and ship production C/C++ features and optimizations in the CUDA driver and runtime to improve latency, throughput, and efficiency for AI, deep learning, and HPC workloads.
Role type
Senior IC systems software engineer (CUDA driver/performance)
Builds
CUDA driver and runtime capabilities, performance optimizations, and production C/C++ features
Domain
AI, deep learning, HPC, GPU computing
Deliverable
production ML models
Required skills
C/C++ systems programming, operating systems and concurrency, computer architecture, performance engineering, root cause analysis, hardware/software co-design
Preferred skills
GPU/accelerator driver development, pre-silicon analysis, performance modeling, Python for data analysis
Technologies
CUDA, C/C++, Linux, GPU architectures
Responsibilities
Design, implement, validate, and ship performance-centric features in the CUDA driver and runtime; Optimize critical execution paths including kernel launch, synchronization, and memory management; Own complex performance problems end-to-end from hypothesis to validation; Establish performance expectations and characterize new silicon; Translate workload evidence into API and programming-model improvements; Partner with cross-functional teams to drive engineering quality and technical direction
Seniority
Senior, hands-on IC with potential for subsystem leadership