Principal Software Engineer

Job Summary

Focused on end-to-end performance for key CSP/hyperscale customers, the full-time Principal Software Engineer will drive performance characterization, gather feedback for optimization, and ensure the reliability of performance tools in a remote or onsite environment.

Key responsibilities

• Drive performance characterization work streams with engineering teams to ensure understanding of platform performance expectations and tuning options
• Gather and synthesize CSP performance feedback to identify gaps and champion optimization priorities within NVIDIA's teams
• Ensure open-source performance tools are updated and validated for the latest NVIDIA systems, enabling reliable baseline measurements

Required qualifications

• 15+ years of experience in systems performance engineering, ideally in GPU/HPC/ML infrastructure, with a BS or MS in Computer Science or related field
• Proficiency in GPU workload profiling using tools like nsight systems and DCGM metrics
• Understanding of distributed training performance dynamics and statistical methods for performance analysis
• Strong data analysis and visualization skills, with a passion for customer performance improvement
• Ability to communicate performance findings to technical audiences and executive leadership