Senior Solutions Architect, Networking, GPU

This is a fully remote position, open to applicants in China.

📋 Description

• Collaborate closely with account managers to grasp customer needs, promote SONiC + Spectrum whitebox switch solutions, and develop technical winning strategies for AI and HPC opportunities.

• Create comprehensive end-to-end architectures for customer proposals, including configurations for GPU servers, storage connectivity, and SONiC-based leaf-spine fabrics utilizing RDMA/RoCE.

• Develop and validate practical demonstrations and proofs of concept by deploying SONiC switches and GPU servers, configuring networking, installing software stacks, and conducting benchmarks.

• Establish performance, scalability, and reliability through detailed technical evaluations.

• Offer support for large-scale production SONiC switch clusters.

• Take ownership of and manage customized SONiC software projects on Spectrum white-box switches.

• Work in partnership with internal engineering, product, and OEM partners to troubleshoot complex issues related to firmware, drivers, OS, routing, and GPU/network performance.

• Implement fixes in active proofs of concept.

• Collaborate with account and OEM partner teams to identify opportunities, design solutions, conduct technical evaluations, and demonstrate value to customers developing large-scale AI and HPC platforms.

• Provide support for production clusters, including monitoring and troubleshooting.

⛳️ Requirements

• Bachelor’s degree in Computer Science, Electrical/Computer Engineering, or equivalent practical experience.

• Over 6 years of experience in data center or cloud infrastructure roles (solutions architect, systems engineer, network engineer) with direct involvement in presales or customer-facing technical tasks.

• Strong background in data center networking for AI workloads: leaf-spine architectures, high-bandwidth/low-latency fabrics, RDMA/RoCE, and ideally InfiniBand.

• Proficient in configuring and troubleshooting networking in both lab and customer settings.

• Practical experience with SONiC: installing and upgrading, configuring interfaces and routing (BGP/EVPN/VXLAN), utilizing monitoring/telemetry tools, and addressing real incidents.

• In-depth understanding of GPU server architecture: CPU/GPU balance, memory bandwidth, PCIe/NVLink topology, storage and NIC placement, and power/cooling considerations at the rack level.

• Hands-on experience in designing, deploying, or managing AI/HPC clusters using GPU-accelerated servers (either on-premises or in the cloud), including actual involvement in sizing, configuration, and performance optimization.

• Exceptional communication and presentation abilities.

• Capable of elucidating complex technical subjects to highly technical engineers as well as non-technical decision makers.

• Proficient in writing clear design documents and proof of concept reports.

• Fluent written and spoken English is mandatory.

🏝️ Benefits

• Competitive salaries.

• Generous benefits package.