NVIDIA - AI Networking Research Student

**About the Role**
Join NVIDIA’s AI Networking team to research and develop cutting-edge solutions for the world’s fastest AI supercomputers. As a student researcher, you’ll design protocols and algorithms that push the boundaries of distributed AI performance while working alongside industry-leading engineers.

**Key Responsibilities**

**High-Performance Networking**
– Research and optimize communication frameworks for massive-scale AI clusters
– Develop low-latency protocols for distributed training workloads
– Co-design hardware/software solutions with chip architecture teams

**Systems Development**
– Implement production-grade networking software in C++/Python
– Benchmark and optimize NCCL/UCX/UCC communication libraries
– Analyze performance bottlenecks in AI/HPC workloads

**Technical Leadership**
– Contribute to open-source networking projects
– Publish research findings and patent innovations
– Work with cross-functional teams on next-gen technologies

**Qualifications**

**Required**
– Current M.Sc./Ph.D. student in:
– Computer Science
– Electrical Engineering
– Related technical field
– Experience with:
– Distributed ML training (PyTorch)
– C++/Python/Linux development
– Computer architecture fundamentals

**Preferred**
– Knowledge of:
– InfiniBand/Ethernet protocols
– NCCL/UCX libraries
– Open-source contributions
– Prior software development experience

**Why NVIDIA**
– Work on frontier AI infrastructure
– Hands-on with breakthrough technologies
– Collaborate with industry pioneers
– Flexible schedule for students