AI/LLM Network Research Intern (High Speed Network) - 2027 Start (PhD)

What you'll do

• Design, optimize, implement, and deploy transport protocols and congestion control algorithms for AI/LLM applications.
• Develop high-performance AI communication frameworks and protocol stacks, including host-network-application co-design optimizations.
• Track emerging technologies and present system innovations in academic papers.

What they ask for

• Currently pursuing a PhD in computer networking or a related technical discipline.
• Experience in at least one area: TCP/RoCEv2 protocols, Socket or verbs network programming, or data-center congestion control algorithms.
• Knowledge of scale-up protocols such as PCIe, NVLink, and UALink and how they differ from scale-out network protocols.
• High-speed network systems experience; proficiency in one or more of C/C++, Python, or Go; and some knowledge of GPU architecture.

Nice to have

• Familiarity with AI training/inference systems and software-hardware co-design; top-tier networking conference publications.
• Experience developing high-performance communication frameworks, including NCCL, MPI, or RPC libraries.

Tools

TCP, RoCEv2, Socket, verbs APIs, PCIe, NVLink, UALink, C/C++

AI focus

Develop networking protocols, congestion control, and communication frameworks to support AI/LLM applications and improve AI/LLM network scalability, reliability, and performance.