Lambda
Staff HPC Network Architect
$284K–$420K
San Jose Office (First St) +1 · Hybrid · Staff · Full-time · Posted Sep 14, 2026
- Location
- San Jose Office (First St)
San Francisco Office (Fremont St) - Experience
- 7+ years; preferably in HPC, AI/ML, or large-scale cloud infrastructure.
About the job
Architects high-performance networks for large-scale GPU clusters, storage systems, and AI cloud platforms.
What you'll do
- Design network topologies and architecture standards for GPU clusters and multi-tenant environments.
- Evaluate network technologies and partner with compute and storage architects on performance and fault tolerance.
- Guide network automation and telemetry strategies; mentor engineers and cross-functional teams.
What they ask for
- Deep expertise with InfiniBand and advanced Ethernet fabrics, including RoCE and RDMA.
- Knowledge of switching, congestion control, QoS, VXLAN/EVPN, and BGP-based fabric design.
- Experience with SDN, APIs, automation, network telemetry, and optical networking.
- Strong communication, leadership, ownership, and ability to work through ambiguity.
Nice to have
- AI workload profiling, NCCL or MPI collective communication, and distributed-training network tuning.
- Exposure to DPU/SmartNIC technologies such as NVIDIA BlueField, or multi-site interconnect design.
Pay
$284K – $420K USD total pay
Tools
InfiniBand HDR/NDR, RoCE, RDMA, PFC, ECN, VXLAN, EVPN, BGP
AI focus
Designs and evaluates networking for AI workloads and GPU clusters; preferred experience includes distributed-training network tuning and AI workload profiling.
Model-written summary · the original posting is authoritative