Research Intern, Efficient Deep Learning – 2027
NVIDIA
Opportunity published from Workday · applications stay with the original hiring source.
About this opportunity
About the opportunity: NVIDIA is seeking a PhD intern to join the Deep Learning Efficiency Research (DLER) team, focusing on efficient deep learning research that creates real-world impact.
The team concentrates on efficient diffusion language models, multimodal generative models, and efficient agentic AI with hybrid inference orchestration across cloud and edge, alongside post-training model optimization, efficient architecture design, adaptive inference, and resource-efficient training.
You will collaborate with a research team that regularly publishes at top computer vision and machine learning venues, contributing to projects that can directly influence NVIDIA products.
Key responsibilities
- Research, design, and implement novel methods for efficient deep learning in diffusion LLMs, multimodal models, or efficient agentic AI
- Develop sampling efficiency, adaptive unmasking, self-speculation, parallel decoding, and multimodal generation pipelines
- Work on hybrid inference orchestration, routing and scheduling policies, expert delegation, and resource-aware agent loops
- Publish original research and collaborate with internal team members, external researchers, and product groups to transfer technology
Qualifications
- Pursuing a Ph.D. in Computer Science/Engineering, Electrical Engineering, or a related field
- Excellent knowledge of the theory and practice of machine learning and deep learning
- Experience with large language models, diffusion language models, multimodal or vision-language models, or agentic systems
- Hands-on experience with large-scale model training including data preparation and model parallelization, such as tensor and pipeline parallelism
- Outstanding research track record with at least one publication at a top-tier conference like ICML, ICLR, NeurIPS, CVPR, or ICCV
- Excellent communication skills
Nice to have
- Parallel programming experience, such as CUDA
- Interest or experience in hybrid cloud-edge inference, orchestration, or adaptive routing
- Background in pruning, quantization, NAS, or efficient backbones
Benefits
- Eligible for standard NVIDIA intern benefits
- Competitive hourly pay based on position, location, year in school, degree, and experience
Job details
- Pay range: 38 USD - 94 USD per hour
- Application deadline: Applications will be accepted at least until October 9, 2026
About the company: NVIDIA is widely considered one of the technology world's most desirable employers, featuring forward-thinking and hardworking teams experiencing rapid growth.
The company is driven by creative and autonomous innovators with a passion for computer architecture and advanced technology.
Additional information
- This posting is for an existing vacancy.
- NVIDIA uses AI tools in its recruiting processes.
Application: UpVagas lists this opportunity from the hiring company or its recruitment platform. The application is completed at the original source.
