Principal Software Engineer – Scale-up Networking (gpu-centric)
HPE (Hewlett Packard Enterprise)
Multiple Locations
Hybrid
Gpu-aware networking design
C/c++ and linux internals expertise
Cuda and rocm gpu programming
Hewlett Packard Enterprise is a global edge-to-cloud company advancing the way people live and work by delivering innovative solutions for digital transformation
Job Summary
Hewlett Packard Enterprise is a global edge-to-cloud company advancing the way people live and work by delivering innovative solutions for digital transformation.
The role involves architecting and delivering GPU-aware scale-up networking solutions, optimizing communication stacks, and leading upstream contributions to open-source projects.
HPE offers comprehensive health and wellbeing benefits, invests in personal and professional development, and fosters an unconditionally inclusive culture.
Matching Summary
Hewlett Packard Enterprise is a global edge-to-cloud company advancing the way people live and work by delivering innovative solutions for digital transformation.
Skills & Requirements
Must-have
GPU-aware networking design
C/C++ and Linux internals expertise
CUDA and ROCm GPU programming
High-performance networking software
MPI and UCX communication stacks
RDMA and PCIe technologies
DMA engine and IOMMU knowledge
Nice-to-have
Open-source HPC/AI library contributions
NIC architecture experience
NUMA and multi-accelerator tuning
Cloud architectures understanding
Cross domain knowledge
Design thinking
Security-first mindset
Key Requirements
10–15+ years experience in high-performance networking or GPU software
Proven architecture and cross-org technical leadership
Experience mentoring senior engineers
Hands-on with CUDA, ROCm, MPI, SHMEM, Libfabric, UCX
Strong knowledge of Linux internals and memory management