About the role#
The DPU team builds core cloud and AI infrastructure. We focus on software-hardware co-design for compute, networking, and storage. This internship offers a chance to contribute to large-scale systems that support cloud and AI platforms. You will work alongside hardware architects and systems engineers to turn research concepts into production-ready solutions.
What you'll need#
- Current enrollment in a PhD program for Computer Science, Computer Engineering, Electrical Engineering, or a related field.
- Availability for a full-time, 12-week internship during Summer 2027.
- Proficiency in C, C++, or Rust with experience in systems-level development and debugging.
- Strong background in Linux systems development.
- Solid understanding of operating systems, computer architecture, networking, or distributed systems.
- Experience in at least one area: software-hardware co-design, computer architecture, distributed storage, high-performance networking, or AI/ML systems.
- A research record through publications, technical reports, or open-source contributions is preferred.
- Familiarity with tools or concepts such as DPDK, RDMA, eBPF, hypervisors, kernel bypass, device virtualization, LLM serving, or performance profiling.
What you'll do#
- Research and develop DPU-based cloud acceleration and AI infrastructure.
- Explore software-hardware co-design using DPUs and GPUs to optimize distributed training and inference.
- Create techniques to accelerate AI training, focusing on communication, resource management, and memory systems.
- Analyze and optimize performance across hardware, device drivers, OS kernels, and user-space runtimes.
- Build and evaluate prototypes using cloud and AI workloads at scale.
- Collaborate on technical proposals, architecture designs, patents, and research publications.
Location & details#
- Location: Seattle, Washington.
- Term: Summer 2027.
- Status: Paid, full-time internship.


