Senior Solutions Architect – CSP System

About the role

As a Senior Solutions Architect focusing on Cloud Service Providers (CSPs) in China, you will be a core technical pillar of NVIDIA’s CSP Solutions Architecture team. You will drive GPU and AI infrastructure strategy, optimize system‑level solutions, and engage with high‑value customers to accelerate large‑scale AI training, inference, and gaming workloads.

Key responsibilities

• Partner with Sales, Business Development and CPM teams to embed NVIDIA GPU and AI technologies into top‑tier Chinese CSP accounts.
• Serve as the primary technical authority for GPU cluster architecture, AI workload deployment, heterogeneous computing tuning, and full‑stack software optimization.
• Unlock CPU‑GPU co‑optimization (Vera) for reinforcement‑learning and Agentic AI workloads, eliminating data‑movement bottlenecks.
• Lead open‑source contributions, deliver upstream patches, and create China‑localized best practices for NVIDIA AI infra stacks.
• Conduct deep GPU workload bottleneck analysis and provide system‑level, kernel‑level, and framework‑level tuning guidelines.
• Act as the liaison between Chinese CSP customers and NVIDIA global engineering, feeding local requirements into product roadmaps while ensuring export‑compliance.
• Run technical workshops, PoCs, and pilot projects to demonstrate business value and drive large‑scale adoption.
• Monitor emerging trends such as Agentic AI, LLM inference optimization, cloud gaming AI, and next‑gen data‑center architectures.
• Mentor junior SA team members and standardize technical engagement processes.

Required profile

• Extensive experience with GPU architecture and AI system optimization.
• Strong background in cluster networking and large‑scale data‑center workloads.
• Proven track record of open‑source AI infrastructure contributions.
• Ability to translate complex technical requirements into scalable solutions for hyperscale CSP environments.

Required skills

• GPU architecture
• AI infrastructure optimization
• Cluster networking
• Open‑source AI stack contributions
• CPU‑GPU co‑optimization (Vera)
• Reinforcement learning workloads
• Agentic AI
• LLM inference optimization
• Cloud gaming AI
• Data‑center system architecture