Description
Hewlett Packard Enterprise is hiring a Principal Engineer to architect and deliver scale-up GPU networking for high-performance computing and AI workloads. The role focuses on GPU-aware networking, GPU-to-NIC data movement, CUDA and ROCm integration, Libfabric and UCX optimization, NUMA and multi-NIC performance tuning, open-source upstream contributions, and debugging across driver, runtime, GPU, kernel, and user-space layers. It is a hybrid position requiring approximately two days per week at an HPE office.
