Principal Software Engineer, Networking - AI Infrastructure Innovation

OCI (Oracle Cloud) AI Infrastructure Innovation team is pioneering the creation of next-generation AI/HPC networking for GPU superclusters at massive scale. Our mission is to design and deliver state‑of‑the‑art RDMA‑based networking—spanning frontend and backend fabrics—that enables customers to achieve high performance for AI training and inference. You will define architecture, lead complex system design, and implement innovative networking software that advances RDMA for GPUs and accelerates storage access. If you thrive at the intersection of large‑scale distributed systems, high‑speed networking, and AI workloads, this role offers the opportunity to push the boundaries of what’s possible. Lead architecture, system design, and implementation for high‑performance RDMA solutions across OCI’s AI/HPC platforms, including frontend and backend fabrics. Innovate on network and TCP performance and identify changes required across (Kernel, NIC, switch, transport, protocol, storage, GPU comms) Develop production‑grade, high‑performance software features with rigorous reliability, observability, and security. Define performance goals and success metrics; design benchmarks and conduct large‑scale experiments to validate throughput, latency, and tail behavior. Collaborate with GPU platform, storage, database, and control‑plane teams to deliver end‑to‑end solutions and influence OCI‑wide network architecture and standards. Mentor engineers, provide technical leadership/reviews, and contribute to long‑term roadmap and technical strategy.

#J-18808-Ljbffr