Cloud Acceleration Research Intern (DPU & AI Infra) - 2027 Start (PhD)

What you'll do

• Develop and evaluate techniques to accelerate distributed AI training and inference, including communication, data movement, resource management, and memory or cache systems.
• Analyze and optimize performance across hardware, drivers, operating-system kernels, communication libraries, and runtimes; build prototypes and test them with representative workloads.
• Collaborate with hardware, systems, and AI infrastructure teams to advance research ideas toward production and contribute to technical proposals, designs, publications, and patents.

What they ask for

• Currently pursuing a PhD in Computer Science, Computer Engineering, Electrical Engineering, or a related technical field.
• Able to commit to a full-time, 12-week internship during Summer 2027.
• Proficiency in C/C++ or Rust, including systems-level development and debugging; strong Linux systems development experience.
• Solid understanding of operating systems, computer architecture, networking, or distributed systems, and a background in at least one listed area: co-design, architecture, distributed storage, high-performance networking, or AI/ML systems.

Nice to have

• Research record shown through publications, technical reports, open-source contributions, or substantial research projects; experience designing, implementing, and evaluating production-quality or research prototype systems.
• Familiarity with DPDK, RDMA, eBPF, high-performance communication libraries, hypervisors, kernel bypass, device virtualization, hardware offload, LLM serving, disaggregated inference, KV-cache management, distributed data loading or storage, performance modeling, profiling, or benchmarking.

Tools

DPDK, RDMA, eBPF, C/C++, Rust, Linux, DPUs, GPUs

AI focus

Research and optimize AI/ML infrastructure, including distributed training and inference, using DPU, GPU, and custom hardware co-design.