Performance Engineer

NVIDIA is seeking a Performance Engineer to join its Perflab team in Shanghai. The successful candidate will work with innovative GPUs and optimize performance throughout the product lifecycle.

Job Summary: As a Performance Engineer at NVIDIA, you will design, configure, and execute benchmark methodologies and real-world use cases to showcase product capabilities across AI/LLM, gaming, workstation, content-creation, and system workloads. You will analyze performance metrics and system telemetry to identify bottlenecks, regressions, and optimization opportunities.

Responsibilities:

• Design, configure, and execute representative benchmark methodologies and real-world use cases
• Analyze performance metrics and system telemetry to identify bottlenecks, regressions, and optimization opportunities
• Produce clear, data-driven competitive analyses and technical reports
• Develop and debug automation scripts for benchmark performance and system monitoring data collection
• Build and maintain scalable automation frameworks, AI agents, dashboards, and data pipelines

Requirements:

• BS or MS degree in Computer Science, Electrical Engineering, Computer Engineering, or a related technical field
• 2+ years of equivalent practical experience
• Strong Python skills or proficiency in another scripting or programming language
• Hands-on interest in AI/LLM workloads, PC gaming, workstation applications, and content-creation workflows
• Solid understanding of PC and server architecture
• Experience working with large datasets and strong analytical and troubleshooting skills

Nice to Have:

• Familiarity with NVIDIA GPUs and NVIDIA Technologies
• Working knowledge of AI inference or training, fine-tuning concepts
• Experience building, benchmarking, or operating AI platforms and LLM inference stacks
• Experience developing AI agents or agentic workflows
• Knowledge of containerized and virtualized workflows, including Docker, Kubernetes, and virtual machines