NVIDIA AI Networking Certification Exam

Test takers will be assessed on their ability to analyze and design robust AI data center architectures, select appropriate network technologies (Ethernet, InfiniBand), differentiate storage options for AI throughput, and optimize topologies for performance, scalability, and redundancy in NVIDIA-powered environments.

• Describe an AI factory networking architecture and its components (e.g., GPUs, NVIDIA BlueField™, scalable units, switches).
• Describe rail-optimized topologies for high-performance AI workloads.
• Describe GPU-to-GPU communications.

5%
Exam Weight

Tasks focus on configuring, optimizing, securing, and troubleshooting NVIDIA Spectrum-X switches. The test taker must demonstrate skills using tools such as NVIDIA NetQ™, CloudAI Benchmark, and DPUs, define advanced quality of service (QoS) and telemetry, monitor and tune network performance, and resolve complex real-world connectivity and microsegmentation scenarios in Ethernet-based AI fabrics.

• Configure NVIDIA Spectrum-X switches for RDMA over Converged Ethernet (RoCE) to enable high-speed, low-latency communication.
• Enable and verify QoS, explicit congestion notification (ECN), priority flow control (PFC), and advanced features like adaptive routing and telemetry
• Configure multi-tenancy Border Gateway Protocol Ethernet VPN (BGP-EVPN) to isolate tenant workloads.
• Use NVIDIA DSX Air to simulate network environments and identify potential issues.
• Diagnose congestion or packet loss using in-band telemetry and NVIDIA What Just Happened™ (WJH) services.
• Use NetQ for real-time network monitoring, including congestion detection and latency measurements.
• Install NVIDIA DOCA™.
• Configure NVIDIA SuperNIC™ functionality for advanced packet processing and congestion control.

30%
Exam Weight

Tasks cover setting up and maintaining InfiniBand switches, configuring reliable, multi-tenant networks, managing subnet managers and fabric partitioning, monitoring performance with tools like NVIDIA Unified Fabric Manager (UFM™), resolving congestion or topology issues, and validating advanced, rail-optimized AI data center designs for scalability and low latency.

• Perform initial configuration and provisioning, including high availability (HA).
• Configure partition keys (PKeys) to ensure secure multi-tenancy in InfiniBand networks.
• Configure QoS and adaptive routing to dynamically adjust paths based on congestion.
• Use UFM to monitor InfiniBand link status and bandwidth utilization.

30%
Exam Weight

The test taker will be assessed on deploying and managing the NVIDIA Network Operator and device plug-ins to enable RDMA/InfiniBand within Kubernetes clusters, validating GPU resources, and ensuring seamless integration with AI workloads. Tasks include troubleshooting cluster-network interfaces, deploying drivers, and scheduling resources on heterogeneous infrastructure.

• Deploy the NVIDIA Network Operator to manage RDMA interfaces and InfiniBand networks within Kubernetes clusters.
• Verify NVIDIA Network Operator functionality.

5%
Exam Weight

Assessment tasks include using diagnostic command-line tools and telemetry (e.g., cl-resource-query, WJH, ib_write_lat, NetQ, UFM) for root-cause analysis of latency, congestion, or connectivity issues in AI networks. The candidate must quickly identify faults in complex environments and recommend effective, production-ready remediation strategies. 

• Use tools like cl-resource-query to check resource allocation in NVIDIA Spectrum-X environments.
• Use WJH services for real-time event analysis.
• Verify low-latency interconnects between GPUs, CPUs, and storage systems.
• Use UFM system health to diagnose InfiniBand issues.
• Use commands like ib_write_lat, ib_write_bw, ibping, ibstat, ibdiagnet, ibnodes, and iblinkinfo to diagnose connectivity issues.

20%
Exam Weight

Assessment tasks focus on automating NVIDIA Spectrum-X and InfiniBand switch deployments using tools like NVUE templates and custom Ansible playbooks, testing for zero-touch operations, and maintaining configuration integrity across multi-vendor, multi-device AI networking environments, including automating RoCE, virtual local area network (VLAN), and high-availability settings.

• Manage NVIDIA Spectrum-X switch configurations through NVUE templates.
• Write Ansible playbooks to automate network setup tasks like VLAN creation or RoCE configuration.

10%
Exam Weight