Senior Manager, Customer Facing Architecture

Overview

• NVIDIA's Spectrum-X Ethernet scale-out GPU networking solution targets AI Clusters as a product.
• ConnectX network adapters, and BlueField DPUs into a unified, AI-optimized fabric that delivers performance, resiliency, debuggability at scale designed with a top-down approach using HW-SW co design methods.
• We are looking for a Team Leader to build and run NVIDIA’s Customer Facing Architects group focused on Spectrum-X in Israel.
• This team serves as the technical connection between NVIDIA's internal AI Networking compose group and our top customers.
• These include hyperscalers, cloud service providers, sovereign AI initiatives, and corporate AI solution developers rs.
• What you’ll be doing: Team Leadership, Build and guide a team of senior Spectrum-X customer-facing architects.
• Set technical direction, own hiring and mentoring, and establish clear roadmap for customer early engagement, onboarding and tracking lifecycle.
• Customer Architecture & Technical Ownership, Serve as the senior technical authority for Spectrum-X deployments at hyperscalers, national AI programs, and enterprise AI factories.
• Drive E2E solution adoption, lead sophisticated technical architecture issues in the solution, and guide customers through configuration, tuning, and performance debugging at the NIC, switch, and fabric level.
• Convey customer's needs, and lead strategic PoCs and benchmarks to demonstrate Spectrum-X advantages.
• Product Feedback, Turn field findings and customer difficulties into concrete feedback for R&D, ASIC, and software teams while partnering with AI Networking Architecture to make sure our deployment realities shape next-gen Spectrum-X design.
• Work with Product Management to influence reference architectures and go-to-market strategy and represent Spectrum-X at customer briefings and industry events such as SC, OCP, and Hot Interconnects What we need to see: BSc/MSc/PhD in Electrical/Computer Engineering, Computer Science, or a related engineering field from leading academic institute.
• 12+ overall years of experience in networking, solutions/field engineering, or network architecture related to data center operations 5+ years in management role.
• Deep, hands-on expertise in data center fabric design: Clos/leaf-spine topologies, BGP, L2/L3 congestion signalling, datapath configurations, and L2/L3 switching at scale.
• Proven understanding of RDMA networking: RoCE v2, InfiniBand semantics, congestion control (ECN/PFC/DCQCN), and the interaction between transport behavior and AI training communication patterns (all-reduce, all-to-all).
• Proven experience with customer-facing technical roles — pre-sales architecture, field engineering, critical issues, or professional services — working with major hyperscalers or CSP accounts.
• Demonstrated ability to lead distributed technical teams, set direction, and deliver results in ambiguous, fast-moving environments.
• Excellent communication skills; ability to present to senior technical and executive collaborators.
• Ways to stand out from the crowd: Direct experience with NVIDIA Spectrum switches, ConnectX NICs, or BlueField DPUs in production or lab environments.
• Familiarity with AI/ML training frameworks (PyTorch, NCCL) and how collective communication patterns work well with network fabric design.
• Background in network telemetry, observability tooling, or adaptive routing algorithms.
• Experience with multi-tenant AI fabric isolation, network slicing, or traffic engineering in GPU cluster environments.
• Prior collaboration with ASIC or silicon teams on network feature definition.