Solution Architecture Intern, AI Infra - 2026

Overview

• NVIDIA has been transforming computer graphics, PC gaming, and accelerated computing for more than 25 years.
• It’s a unique legacy of innovation that’s fueled by great technology—and amazing people.
• Today, we’re tapping into the unlimited potential of AI to define the next era of computing.
• An era in which our GPU acts as the brains of computers, robots, and self-driving cars that can understand the world.
• Doing what’s never been done before takes vision, innovation, and the world’s best talent.
• As an NVIDIAN, you’ll be immersed in a diverse, supportive environment where everyone is inspired to do their best work.
• Come join the team and see how you can make a lasting impact on the world.
• Join NVIDIA, a groundbreaking leader in AI computing and visual technologies, at the forefront of innovation.
• As an AI in Industry Solution Architecture Intern, you'll be integral to our mission of redefining industries through AI and HPC.
• Our Solution Architect team builds innovative AI computing platforms, analyzes applications, and delivers outstanding value to our customers.
• This role offers a remarkable opportunity to harness NVIDIA's newest technologies to optimize large models, develop sophisticated AI workflows, and empower our clients with advanced AI solutions.
• What you will be doing: Develop and gain in‑depth understanding of open‑source inference frameworks such as SGLang and vLLM; collaborate with the community on new features and operators, performance optimization, and model enablement.
• Design and implement CUDA kernels/operators (e.g., GEMM, attention and related primitives) for efficient and scalable LLM inference and training.
• Develop and optimize KV‑cache offloading frameworks for LLM scenarios, enabling multi‑level KV‑cache offloading and reuse on CPU/SSD/remote storage to accelerate inference (team project: https://github.com/taco-project/FlexKV).
• Take ownership of R&D work related to compute performance in distributed training, continuously exploring methods and techniques for performance optimization.
• Conduct in‑depth research on computational problems in machine learning, summarize common computational patterns and requirements, and develop sample code, acceleration libraries, or framework components.
• What we need to see: Pursuing a Bachelor or Master or PhD in Electrical Engineering, Automation, Computer Science, Computational Mathematics, or related fields.
• Strong interest in accelerated computing, parallel computing, and heterogeneous computing, and willingness to dive deep into these areas.
• Solid programming skills; good understanding of data structures and general concepts of computer systems.
• Strong ability to learn and adapt, with good skills in analyzing and formulating problems and exploring solutions independently.
• Ways to stand out from the crowd: Familiarity with heterogeneous computing, distributed training, parallel computing, or other high‑performance computing areas.
• Experience in performance analysis, performance modeling, or performance optimization, and contributions to open‑source frameworks.
• Strong capability in defining new problems and exploring solution spaces; this is critical for the role.
• Proficiency with AI‑assisted programming tools.