All 6 domains
24 official objectives
84–122 hrs suggested study
-
Domain 01
5%
Domain 1 explains how an NVIDIA AI factory network is put together: the GPUs, BlueField devices, scalable units and switches, why rail-optimized topologies exist, and how GPUs exchange data inside and across servers.
Inside this domain
- 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
Watch for NVLink spans the cluster
- Objectives
- 3
Read the domain guide →
-
Domain 02
30%
The joint-largest NCP-AIN domain covers the Spectrum-X Ethernet platform: lossless RoCE, congestion control and adaptive routing, multi-tenant EVPN, simulation in NVIDIA Air, telemetry with WJH and NetQ, DOCA installation and SuperNIC setup.
Inside this domain
- 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), and priority flow control (PFC), as well as advanced features like adaptive routing and telemetry
- Configure multi-tenancy Border Gateway Protocol Ethernet VPN (BGP-EVPN) to isolate tenant workloads
- Use NVIDIA Air to simulate network environments and identify potential issues
- and 4 more objectives
Watch for PFC solves congestion
- Objectives
- 8
Read the domain guide →
-
Domain 03
30%
InfiniBand Networking shares the top weight on NCP-AIN. It covers bringing up an InfiniBand fabric with a highly available subnet manager, isolating tenants with partition keys, tuning QoS and adaptive routing, and monitoring links with UFM.
Inside this domain
- 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
Watch for Cables are the problem
- Objectives
- 4
Read the domain guide →
-
Domain 04
5%
This domain covers bringing NVIDIA networking into Kubernetes: deploying the NVIDIA Network Operator to manage RDMA interfaces and InfiniBand networks for pods, and verifying that drivers, device plugins and secondary networks work.
Inside this domain
- Deploy the NVIDIA Network Operator to manage RDMA interfaces and InfiniBand networks within Kubernetes clusters
- Verify NVIDIA Network Operator functionality
Watch for Default CNI carries RDMA
- Objectives
- 2
Read the domain guide →
-
Domain 05
20%
Troubleshooting Tools is the third-largest NCP-AIN domain. It tests whether you can pick and read the right tool: cl-resource-query and WJH on Spectrum-X, interconnect checks across GPUs, CPUs and storage, UFM system health and the InfiniBand command set.
Inside this domain
- Use tools like cl-resource-query to check resource allocation in 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
- and 1 more objectives
Watch for ibping just works
- Objectives
- 5
Read the domain guide →
-
Domain 06
10%
Automation and Configuration covers managing Spectrum-X switches as code: using NVUE's object model, configuration files and templates, and writing Ansible playbooks that create VLANs or apply RoCE settings repeatably across the fabric.
Inside this domain
- Manage Spectrum-X switch configurations through NVUE templates
- Write Ansible playbooks to automate network setup tasks like VLAN creation or RoCE configuration
Watch for Set means live
- Objectives
- 2
Read the domain guide →
Domain names, weights and objectives come from NVIDIA's official outline (NCP-AIN exam study guide (MAY26 revision, current)). Question counts and study hours are editorial planning estimates, not official figures. Last reviewed October 2026.