Senior Software and System Architect

Overview

• NVIDIA is looking for an outstanding Senior Networking Software Architect to join the NIC/DPU Software and Firmware Architecture group.
• In this role, you will help define the next generation of NVIDIA datacenter and AI networking platforms, with focus on DPU management, QoS, performance, telemetry, and software architecture across stacks.
• You will work closely with hardware designers, firmware/kernel driver teams, system engineers, validation, product management and customers.
• The role spans early architecture definition, pre-silicon design, bring-up, and production readiness for large-scale AI and cloud datacenter deployments.
• What You’ll Be Doing: Own software and system architecture for next-generation DPU management, QoS, performance, telemetry, and observability features.
• Define end-to-end control and management flows across DOCA, host drivers, embedded firmware, BMC, management controllers and external management systems.
• Specify telemetry and observability requirements, including counters, logs, traces, events, health monitoring, debug data, and streaming telemetry.
• Define management interfaces and APIs for configuration, provisioning, lifecycle operations, diagnostics, and field serviceability.
• Write clear architecture specifications, interface definitions, flow diagrams, and design documents for software, firmware, and system teams.
• Partner with R&D teams to translate high-level architecture into implementable designs and guide features through development, validation, silicon bring-up, and production.
• Analyze system performance bottlenecks, interoperability issues, telemetry gaps, and customer-reported issues, then feed learnings into future architecture.
• Collaborate with system and cluster architects to ensure NIC/DPU features fit end-to-end AI datacenter and cloud networking designs.
• What We Need To See: B.Sc.
• in Computer Engineering, Computer Science, Electrical Engineering, or equivalent experience.
• 9+ years of experience in networking, system software, embedded software, firmware, or datacenter infrastructure.
• Proven experience in software architecture, or technical leadership roles.
• Deep understanding of networking concepts and protocols such as Ethernet, TCP/IP, RDMA/RoCE, congestion control, QoS, virtualization overlays, and traffic management.
• Strong background with DPUs, SmartNICs, or other high-performance networking devices.
• Experience with system management, provisioning, monitoring, telemetry, diagnostics, or lifecycle-management flows.
• Familiarity with management protocols and frameworks such as Redfish, PLDM, MCTP, IPMI, gNMI, SNMP, Netconf, REST, or gRPC-based APIs.
• Ability to lead cross-functional architecture discussions across software, firmware, hardware, validation, product, and customer-facing teams.
• Excellent written and verbal communication skills, including the ability to create clear architecture documents and present trade-offs.
• Ways To Stand Out From The Crowd: Experience defining software architecture for DPU products, including management, telemetry, QoS, performance, security, virtualization, or offload features.
• Hands-on background with Linux networking, device drivers, firmware, embedded Linux, BMC software, DOCA, DPDK, OVS or Kubernetes networking.
• Experience with performance counters, profiling tools, eBPF, Prometheus, Grafana, dashboards, heat maps, or large-scale telemetry systems.
• Experience in defining and developing GAI-based analysis tools to extract insights from telemetry data and streams.
• Background in RAS, diagnosability, serviceability, field failure analysis, production debug, or customer escalation handling.
• NVIDIA is widely considered one of the technology world’s most desirable employers.
• We have some of the most forward-thinking and hardworking people in the industry working with us, and we are building the networking platforms that power the AI revolution.
• If you are passionate about NICs, DPUs, system software, and large-scale AI networking, and you want to shape the management, telemetry, QoS, and performance architecture of future NVIDIA platforms, we want to hear from you.