Senior Solutions Architect, Networking and GPU System

Role overview Design and validate networking and GPU server solutions for large-scale AI and high-performance computing environments. The role combines customer-facing architecture work, hands-on proofs of concept, and support for production clusters, with collaboration across account, engineering, product, and hardware partners.

Responsibilities - Translate customer requirements into end-to-end architectures spanning GPU servers, storage connectivity, and leaf-spine networks using RDMA/RoCE. - Work with account teams to assess opportunities and develop technical approaches for AI and HPC deployments. - Build and run demonstrations and proofs of concept, including switch and server setup, software installation, network configuration, and performance testing. - Support production network clusters and troubleshoot issues across firmware, drivers, operating systems, routing, and GPU or network performance. - Manage custom network operating system projects on white-box switches and bring fixes into active evaluations. - Prepare design documents and evaluation reports, explaining technical findings to engineers and decision-makers.

Requirements - Bachelor’s degree in computer science, electrical or computer engineering, or equivalent practical experience. - At least six years in data center or cloud infrastructure, including customer-facing technical or presales work. - Strong experience with data center network design and troubleshooting, including leaf-spine fabrics and RDMA/RoCE; InfiniBand experience is useful. - Practical experience installing, upgrading, configuring, monitoring, and troubleshooting SONiC, including routing such as BGP, EVPN, or VXLAN. - Working knowledge of GPU server architecture, including compute balance, memory bandwidth, interconnect topology, storage and NIC placement, and rack power and cooling. - Experience sizing, deploying, operating, or tuning GPU-accelerated AI or HPC clusters; fluent written and spoken English.

Nice to have - Contributions to open-source networking or AI infrastructure, or published technical talks or papers. - Coding experience with SONiC, RDMA applications, collective communication libraries, or distributed training systems. - Large-scale SONiC deployments integrated with GPU clusters and RDMA fabrics; familiarity with commercial data center networking platforms.