PhD Research Intern, Nemotron India – 1 Year Intern 2027

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.   This internship focuses on advancing the algorithms and model architectures behind language foundation models such as Nemotron, pushing model quality across reasoning, agentic, and multi-lingual specifically focused on India-specific outcomes. The work is model-improvement first, centered on designing new architectural components, training objectives, and efficiency techniques that scale. Progress is driven through disciplined experimentation: targeted ablations, strong diagnostics, and repeatable side-by-side quality evaluation. Collaboration spans researchers and engineers across multiple teams to translate promising ideas into measurable, reproducible gains.   What you’ll be doing: Research, implement, and validate model architecture and algorithm changes that improve Nemotron, especially for India. Explore and prototype improvements across language understanding, voice agents, translation, and token efficiency (distillation / compression). Improve training and inference efficiency through architectural and post-training techniques (e.g., compute/memory optimizations, distillation or compression strategies) while preserving or improving quality. Design data processing and curation workflows that raise training set quality for Indic language capabilities, including using strong off-the-shelf models for filtering, labeling, and quality assessment.   What we need to see: Pursuing PhD in Computer Science, Computer Engineering, or a closely related field. Research or project experience in LLMs, and multi-lingual data. Strong Python and PyTorch proficiency proven through research code bases, publications, or substantial projects. Experience with at least one of the following: model training, post-training/fine-tuning, dataset curation, or benchmarking/evaluation pipelines for large language models. Comfortable working in Linux and using standard software development practices for ML research code (version control, testing, experiment reproducibility).   Ways to stand out from the crowd: Interested in or experienced with Indic foundation models and challenges with multi-lingual data and evals. Exposure to large-scale training or performance-sensitive experimentation (multi-GPU / multi-node, distributed data loading, performance profiling).   Join us to help build the next generation of Nemotron for India! NVIDIA is committed to fostering a diverse work environment and proud to be an equal opportunity employer. As we highly value diversity in our current and future employees, we do not discriminate (including in our hiring and promotion practices) on the basis of race, religion, color, national origin, gender, gender expression, sexual orientation, age, marital status, veteran status, disability status or any other characteristic protected by law.