Deep Learning Kernel Software Performance Architect

NVIDIA is seeking a Deep Learning Kernel Software Performance Architect to optimize GPU kernel performance for state-of-the-art data-center platforms.

You will build automated, data-driven workflows to detect, explain, and prevent performance regressions across key deep learning workloads, partnering closely with kernel developers, compiler teams, infrastructure, and architecture/performance groups.

Responsibilities:

• Performance analysis, optimization, and debugging:
• Analyze performance of GPU-accelerated kernels and key deep learning building blocks
• Identify gaps with baselines or projections and optimize kernel performance to fill gaps
• Debug performance issues end-to-end
• Automation and regression infrastructure (Python-heavy):
• Develop and maintain Python-based automation for performance testing and analysis
• Design and operate performance test workflows
• Cross-team collaboration and operating model:
• Work with kernel developers and compiler teams to ensure performance checks are practical and scalable
• Collaborate with chip architecture and modeling teams to solidify performance methodology
• Partner with SWQA and infrastructure teams for execution at scale and reliable pipelines/dashboards

Requirements:

• Master's or Ph.D. degree or equivalent experience in Computer Science, Computer Engineering, Applied Math, or related field
• Strong programming ability in Python and C/C++ with 2+ years of working experience
• Solid fundamentals in computer architecture, parallel programming, and performance reasoning
• Experience with performance analysis workflows
• Comfortable working across teams and driving issues to decision/closure with clear communication

Nice to Have:

• Experience with high-performance kernels or math libraries
• GPU programming and performance experience
• Strong ML/DL workload understanding
• Familiarity with simulators or analytical modeling