Senior Performance Architect - Heterogeneous Workload Optimization

NVIDIA, a leader in computer graphics and AI computing, is seeking a Senior Performance Architect to optimize heterogeneous workloads. You will architect and maintain custom profiling frameworks for CPU and GPU environments, conduct benchmarking, and develop predictive models for hardware procurement.

Key Responsibilities:

• Design and maintain unified profiling frameworks for CPU and GPU environments.
• Conduct in-depth benchmarking of EDA applications to analyze memory access patterns and instruction-level parallelism.
• Use GPU profilers to identify inefficiencies such as warp divergence and PCIe/NVLink bottlenecks.
• Develop tools to monitor and optimize high-watermark memory usage in multi-terabyte EDA builds.
• Create predictive models to guide hardware procurement and cloud instance selection.

Requirements:

• Experience with CUDA programming and GPU profiling tools like NVIDIA Nsight Systems/Compute.
• Knowledge of profiling tools such as perf, eBPF, VTune, or Valgrind.
• Strong benchmarking skills and ability to analyze complex performance data.
• Experience with distributed compute environments like Slurm, LSF, or Kubernetes.
• BS, MS, or PhD in Computer Science, Electrical Engineering, or related field with 8+ years of relevant experience.

NVIDIA offers competitive salaries and a comprehensive benefits package, including equity.