NVIDIA NCP-AII Questions & Answers

The option would apply to “A migration swaps two similarly labeled fabric cables and only one of four MPI ranks loses inter-node connectivity”. The direct fix is Use a per-port peer discovery and rank-to-node cable map to isolate the incorrect physical connection. This question instead requires “Reconcile breakout cable legs, configured port modes and server interfaces against the reference topology”.

The related but distinct situation is “A fabric acceptance team sees all links up but does not check whether storage ports were accidentally wired to GPU collective switches”. Therefore Use a physical topology reconciliation with port identifiers and approved network role assignments. For the stated problem, “Reconcile breakout cable legs, configured port modes and server interfaces against the reference topology” addresses the central condition.

This option serves the scenario “A rack diagram claims both InfiniBand rails are independent, but an LLDP or switch inventory shows one server connected twice to rail A”. The appropriate intervention is Trace each cable endpoint against the approved rail map and correct the crossed connection. The present scenario is distinguished by “Reconcile breakout cable legs, configured port modes and server interfaces against the reference topology”.

The related but distinct situation is “A GPU node's driver initializes but the cluster image lacks the libraries used by the production training framework”. The decisive evidence is that A small framework GPU operation and collective test run successfully from the approved image. The deciding requirement here is “A prolonged load run maintains permitted temperatures, clocks and application throughput without faults”.

This would be a meaningful test for another issue (A server lists all GPUs successfully but distributed training fails whenever an application uses inter-GPU communication), but not for the required outcome: A prolonged load run maintains permitted temperatures, clocks and application throughput without faults.

The proposed validation does not isolate the reported fault. An effective closure test requires evidence that A prolonged load run maintains permitted temperatures, clocks and application throughput without faults.

This option serves the scenario “A host updates the CUDA toolkit but leaves an incompatible NVIDIA kernel driver and the application fails to initialize CUDA”. Therefore Align the installed GPU driver and userspace software with NVIDIA's supported compatibility information. The present scenario is distinguished by “Install and sign the required driver module using approved key enrollment procedures”.

The suggested action targets another problem where Removing GPU control software from an active compute node can interrupt every GPU process. It does not directly address this case, which needs Install and sign the required driver module using approved key enrollment procedures.

The suggested action targets another problem where Installed driver packages and the currently loaded kernel module can be different versions. It does not directly address this case, which needs Install and sign the required driver module using approved key enrollment procedures.

The option would apply to “A dual-rail rack has all high-speed ports connected but two cables are accidentally crossed between fabrics”. The operational evidence follows from the mechanism that Correct link state does not guarantee that each cable terminates on the correct rail or security domain. The present scenario is distinguished by “Matching connector geometry does not imply optical signaling compatibility between transceiver endpoints”.

The related but distinct situation is “An installation map labels AOC and DAC assemblies interchangeably and technicians order the wrong reach for distant racks”. This case illustrates that Active optical and direct-attach copper cables differ in reach, power and handling requirements. The deciding requirement here is “Matching connector geometry does not imply optical signaling compatibility between transceiver endpoints”.

This option serves the scenario “A 400G fabric port uses a short passive copper cable beyond its approved reach and link quality is unstable”. The operational evidence follows from the mechanism that Passive copper, active copper and optical transceivers have distinct reach and signal-integrity envelopes. This question instead requires “Matching connector geometry does not imply optical signaling compatibility between transceiver endpoints”.

The option would apply to “An engineer updates GPU board firmware outside the planned change window while the cluster scheduler continues submitting training jobs”. Without correction, Uncoordinated firmware changes can terminate jobs and produce untraceable performance regressions. This question instead requires “Broad restart during a failed flash can turn a contained device issue into rack-wide downtime”.

The related but distinct situation is “An AI factory firmware update changes BMC inventory fields and a monitoring dashboard now reports healthy GPUs as missing”. Without correction, A broken management schema can hide real component faults under apparently successful patching. For the stated problem, “Broad restart during a failed flash can turn a contained device issue into rack-wide downtime” addresses the central condition.

This option serves the scenario “An HGX server cluster has mixed GPU and BMC firmware revisions after a partial maintenance window, but the team runs burn-in as if versions match”. Without correction, Mixed unsupported firmware levels can cause intermittent device initialization and fabric instability. The present scenario is distinguished by “Broad restart during a failed flash can turn a contained device issue into rack-wide downtime”.

This option serves the scenario “A GPU server's DPU has an approved high-speed physical link but no functioning OOB path for emergency firmware recovery”. The decisive evidence is that A simulated host outage still permits authenticated DPU status retrieval and approved recovery actions. This question instead requires “Version inventories from host and Arm agree with the validated release matrix and network acceleration tests succeed”.

This alternative addresses the separate case “A BlueField-3 Arm-side service is stopped during maintenance, after which the host's accelerated network path fails to initialize”. A completed remediation must demonstrate that Arm-side health and host-side interface state recover in the expected order with working test traffic. For the stated problem, “Version inventories from host and Arm agree with the validated release matrix and network acceleration tests succeed” addresses the central condition.

The option would apply to “A BlueField DPU is operating in NIC mode, but the security architect expects the Arm control plane to enforce independent host network policy”. The strongest acceptance result is that An inspected mode report confirms embedded ownership and a policy test denies an unauthorized host flow, not merely that a command executed. The present scenario is distinguished by “Version inventories from host and Arm agree with the validated release matrix and network acceleration tests succeed”.

This alternative addresses the separate case “A high-speed optical link shows intermittent FEC corrections rising under rack vibration despite no link-down alarms”. The operational evidence follows from the mechanism that Corrected FEC events can reveal deteriorating signal margin before hard link failures occur. The deciding requirement here is “Passive copper, active copper and optical transceivers have distinct reach and signal-integrity envelopes”.

The option would apply to “An engineer installs an optic that matches the physical cage but uses a wavelength incompatible with the remote transceiver”. The operational evidence follows from the mechanism that Matching connector geometry does not imply optical signaling compatibility between transceiver endpoints. This question instead requires “Passive copper, active copper and optical transceivers have distinct reach and signal-integrity envelopes”.

The related but distinct situation is “An installation map labels AOC and DAC assemblies interchangeably and technicians order the wrong reach for distant racks”. This case illustrates that Active optical and direct-attach copper cables differ in reach, power and handling requirements. For the stated problem, “Passive copper, active copper and optical transceivers have distinct reach and signal-integrity envelopes” addresses the central condition.

This alternative addresses the separate case “A backup head node has no configured remote power-fencing capability, but the team expects automatic recovery from any loss of heartbeat”. The decisive evidence is that A planned failure demonstration follows the documented automatic or manual activation and fencing sequence. For the stated problem, “An HA status check identifies the secondary as active with services working after a controlled takeover” addresses the central condition.

This is weaker evidence because BCM primary and secondary are labels, whereas active service ownership changes when a validated failover occurs. The independent test must show that An HA status check identifies the secondary as active with services working after a controlled takeover.

The candidate evidence checks Configure supported HA fencing, quorum and shared-resource ownership protections before approving automated failover in another context. Here the specific acceptance result is An HA status check identifies the secondary as active with services working after a controlled takeover.

The option would apply to “An operations team uses a single successful training run as evidence that six racks meet the contractual availability target”. The decisive evidence is that A traceable acceptance report includes every node, rack, network path, burn-in duration, exception and signoff. This question instead requires “Asset inventory ties chassis, GPUs, NICs, PSU identities and management firmware to approved bill of materials”.

This is weaker evidence because Hardware chain of custody and baseline versions must be established before subsequent configuration can be compared reliably. The independent test must show that Asset inventory ties chassis, GPUs, NICs, PSU identities and management firmware to approved bill of materials.

This alternative addresses the separate case “A data center imports server settings from a reference rack without checking whether the new deployment uses different network and storage attachments”. The decisive evidence is that Commissioning records reconcile each port and mount with the approved physical design and a working target workload. The deciding requirement here is “Asset inventory ties chassis, GPUs, NICs, PSU identities and management firmware to approved bill of materials”.

This option serves the scenario “A new Intel server uses a BIOS power-saving policy that holds CPU frequency low while the GPU framework needs sustained preprocessing throughput”. The direct failure mode is that Unreviewed power-saving defaults can make an expensive GPU node look underutilized. This question instead requires “Poor NUMA placement can starve otherwise healthy accelerators and reduce training throughput”.

This alternative addresses the separate case “An engineer enables every possible CPU thread for data loading without checking memory locality and context-switch overhead on a multi-socket host”. The primary exposure is that Oversubscribed host threads can slow training and introduce intermittent batch starvation. For the stated problem, “Poor NUMA placement can starve otherwise healthy accelerators and reduce training throughput” addresses the central condition.

The option would apply to “A deployment moves from an Intel to an AMD host while preserving all x86 CPU tuning flags without verifying the new platform topology”. The direct failure mode is that Blind tuning reuse can degrade performance or disable platform-specific reliability protections. The present scenario is distinguished by “Poor NUMA placement can starve otherwise healthy accelerators and reduce training throughput”.

This option serves the scenario “An AI inference server performs correctly at one request per second but misses latency targets when all accelerators share the CPU and network paths”. Without correction, Single-request smoke tests can conceal contention and poor placement under production demand. The present scenario is distinguished by “An undetected inter-GPU fault can make a nominally healthy compute node unusable for training”.

This alternative addresses the separate case “A node passes short GPU tests but its application intermittently throttles after twenty minutes of sustained work”. Without correction, Short acceptance tests can conceal thermally constrained GPUs that disrupt production jobs. The deciding requirement here is “An undetected inter-GPU fault can make a nominally healthy compute node unusable for training”.

The option would apply to “A GPU server passes nvidia-smi query but CUDA sample binaries fail to allocate device memory under an ordinary user account”. Without correction, A scheduler may dispatch jobs to a node that cannot run real tenant workloads. This question instead requires “An undetected inter-GPU fault can make a nominally healthy compute node unusable for training”.