Shi Jin
Senior SDE at AWS Annapurna Labs
PhD in Physics, UW
AWS Annapurna Labs
University of Washington
Seattle, WA
About
I am a Senior Software Development Engineer at AWS Annapurna Labs, working on high-performance communication libraries built on top of the AWS Elastic Fabric Adapter (EFA) — a custom RDMA-capable NIC designed for tightly-coupled HPC and large-scale ML workloads.
My primary focus is libfabric (OFI), the open-source network fabric API that provides portable access to high-speed transports including EFA, InfiniBand, and RoCE. I also contribute significantly to Open MPI for HPC collective communication. Occasionally I collaborate on related projects such as aws-ofi-nccl, NIXL, and UCCL.
Before joining AWS, I obtained a PhD in physics from the University of Washington. During my graduate years, working with Prof. Aurel Bulgac, I studied quantum many-body problems in nuclear physics. My research primarily involved the application of density functional theory (DFT) and its time-dependent extension (TDDFT) on superfluid many-fermion systems — from static nuclear structure to nuclear fission and reactions.
I have expertise in high-performance computing, including MPI, OpenMP, and CUDA C programming, and a solid background in applied mathematics, especially the numerical solution of partial differential equations (PDEs).
I was born in 1991 in Wuhu, Anhui, China. I received a B.S. in physics from the University of Science and Technology of China (USTC) in 2013 before joining the graduate program at the University of Washington.
Open Source Projects
Open Fabric Interfaces — a high-performance networking API providing portable access to RDMA transports including AWS EFA, InfiniBand, and RoCE. My main open source project.
The standard open-source MPI implementation for HPC. I contribute to collective communication and EFA transport support.
A plugin bridging NCCL and libfabric, enabling distributed deep learning training on AWS EFA.