Hands-on experience with mellanox connectx-7 tools
Intel DCG China is seeking a Research Intern to contribute to innovative solutions for disaggregated AI supernode architectures, focusing on system performance and cost optimization. The ideal candidate should be a master's student in a relevant engineering field with specific technical skills, particularly in GPU interconnect protocols and memory pooling
Job Summary
The role focuses on exploring architectural innovations for disaggregated AI supernode designs to optimize cost and scalability.
Candidates will research and prototype Ethernet-native GPU interconnect protocols and distributed memory pooling mechanisms.
This is an on-site internship position located in Shanghai, China, requiring a Master's degree candidate.
Matching Summary
Match Score: 75
Intel DCG China is seeking a Research Intern to contribute to innovative solutions for disaggregated AI supernode architectures, focusing on system performance and cost optimization. The ideal candidate should be a master's student in a relevant engineering field with specific technical skills, particularly in GPU interconnect protocols and memory pooling.
Skills & Requirements
Must-have
Master's student in Engineering or CS
Familiarity with core RDMA operations
Hands-on experience with Mellanox ConnectX-7 tools
Working knowledge of LLM inference benchmarking
Nice-to-have
Passionate and self-motivated attitude
Experience with vLLM or lm-evaluation-harness frameworks
Understanding of system-level performance trade-offs