Performance Modeling Architect

About the role

Join NVIDIA's networking team to shape, optimize, and explore next‑generation networking chips and solutions. You will work cross‑disciplinarily with hardware, software, and research groups to define architectural directions for high‑performance AI and HPC systems.

Key responsibilities

• Develop C++ simulation models to evaluate architectural options for future networking products.
• Analyze communication patterns of deep‑learning, generative AI, inference, and training workloads.
• Identify and resolve performance bottlenecks, balancing trade‑offs across hardware and software.
• Lead technical workgroups and collaborate with architecture, research, hardware, and software teams.
• Explore innovative ideas to improve chip‑level performance and system efficiency.

Required profile

• BSc/MSc in Computer Science, Electrical Engineering, or a related field from a recognized university.
• Strong verbal and written communication skills.
• High ownership mindset with the ability to drive solutions and initiatives.
• 3+ years of hands‑on experience in performance modeling or related domains.
• Proven debugging capabilities.

Required skills

• C++ programming
• Python scripting
• NVLink technology
• Ethernet networking
• InfiniBand (IB) technology