NVIDIA NCP-AII Questions & Answers

This option serves the scenario “After reboot, a previously assigned MIG layout disappears and the scheduler tries to place jobs on nonexistent device UUIDs”. The decisive evidence is that A controlled reboot restores the intended GPU and compute instances before accepting tenant jobs. The present scenario is distinguished by “A concurrent workload test shows distinct assigned MIG identifiers and authorized resource visibility”.

This alternative addresses the separate case “A new multi-tenant inference service assigns a whole GPU to every tiny model despite approved MIG-capable hardware”. A completed remediation must demonstrate that Tenant benchmark runs use only their allocated profiles with expected per-instance resource limits. The deciding requirement here is “A concurrent workload test shows distinct assigned MIG identifiers and authorized resource visibility”.

The option would apply to “An administrator enables MIG mode on a supported GPU but creates no GPU or compute instances and the workload sees no device”. The strongest acceptance result is that The instance inventory lists valid GPU and compute instances and a sample job uses the assigned device, not merely that a command executed. This question instead requires “A concurrent workload test shows distinct assigned MIG identifiers and authorized resource visibility”.

This alternative addresses the separate case “A node's BIOS reports a NUMA topology different from the approved design after a firmware reset”. Without correction, Undetected topology drift can cause large performance variance across nominally identical servers. For the stated problem, “An undetected missing fabric interface can later break one rail or limit all-reduce throughput” addresses the central condition.

The option would apply to “An installer forgets one GPU power connector after servicing a PCIe server and the GPU appears intermittently under load”. Without correction, A partially powered accelerator can disappear or fail when training reaches peak load. The present scenario is distinguished by “An undetected missing fabric interface can later break one rail or limit all-reduce throughput”.

This option serves the scenario “A server reports ECC correctable errors growing quickly during a memory stress test, yet the operator checks only whether the job exits zero”. The primary exposure is that Ignoring climbing corrected errors can allow a latent GPU or memory fault to become an uncorrectable failure. This question instead requires “An undetected missing fabric interface can later break one rail or limit all-reduce throughput”.

That observation does not establish operating effectiveness of Persist the approved NVIDIA runtime configuration and verify daemon startup through reboot. The needed evidence is that A post-reboot Docker test runs an approved CUDA image with the expected GPU devices.

This option serves the scenario “A security baseline blocks device pass-through entirely even though authorized containers require selected GPU access”. A completed remediation must demonstrate that An authorized container sees its allowed GPU UUID while an unauthorized container sees none. The present scenario is distinguished by “A post-reboot Docker test runs an approved CUDA image with the expected GPU devices”.

This alternative addresses the separate case “An administrator installs the NVIDIA Container Toolkit but Docker workloads still show no GPUs because the runtime is not configured”. The strongest acceptance result is that A supported GPU-enabled container lists the expected devices after runtime configuration, not merely that a command executed. The deciding requirement here is “A post-reboot Docker test runs an approved CUDA image with the expected GPU devices”.

The related but distinct situation is “An engineering team sizes GPU network capacity from model checkpoint file size while ignoring synchronization frequency and collective patterns”. The appropriate intervention is Derive fabric demand from observed all-reduce volumes, batch schedule, node count, and expected scaling efficiency. The deciding requirement here is “Create explicit management and data-plane isolation with suitable routing, QoS and access boundaries”.

This alternative addresses the separate case “A storage and training network both connect to the same leaf pair, but the architect ignores competing traffic during checkpoint bursts”. The direct fix is Define traffic classes and capacity headroom for storage checkpoint writes and synchronous GPU collectives. For the stated problem, “Create explicit management and data-plane isolation with suitable routing, QoS and access boundaries” addresses the central condition.

The option would apply to “A new AI factory adds two leaf switches but places their uplinks through a single shared spine and claims full fabric availability”. Therefore Model spine and leaf failures and provide redundant forwarding resources matched to the target recovery level. The present scenario is distinguished by “Create explicit management and data-plane isolation with suitable routing, QoS and access boundaries”.

This option serves the scenario “A cluster completes a single multi-node NCCL run but times out after several hours of continuous all-reduce burn-in”. The key distinction is that Intermittent hardware or congestion faults may emerge only during sustained multi-node communication. The present scenario is distinguished by “Collective failure probability and network behavior change with participant count and communication topology”.

The option would apply to “A collective burn-in reports consistent success but one rack contributes very low bus bandwidth and extends global completion time”. This case illustrates that Synchronous collectives are constrained by slow participants and imbalanced network paths. This question instead requires “Collective failure probability and network behavior change with participant count and communication topology”.

The related but distinct situation is “An NCCL burn-in is interrupted by a scheduled switch update and the report falsely labels the hardware rack as failing”. The controlling explanation is that Burn-in failure interpretation requires knowledge of external disruptions to avoid false hardware diagnoses. For the stated problem, “Collective failure probability and network behavior change with participant count and communication topology” addresses the central condition.

The option would apply to “A newly deployed node passes a CPU-only stress test but its GPU accelerators fail when exposed to simultaneous high-memory loads”. A completed remediation must demonstrate that Simultaneous GPU memory tests complete within errors and thermal limits across all installed devices. The present scenario is distinguished by “Repeated cold boot and sustained compute tests detect every installed GPU without power fault events”.

The candidate evidence checks Collect PCIe link speed, width, device identity and error state alongside normal host inventory in another context. Here the specific acceptance result is Repeated cold boot and sustained compute tests detect every installed GPU without power fault events.

This option serves the scenario “A node's BIOS reports a NUMA topology different from the approved design after a firmware reset”. The decisive evidence is that Topology reports and controlled bandwidth tests align with the approved baseline after configuration correction. This question instead requires “Repeated cold boot and sustained compute tests detect every installed GPU without power fault events”.

The option would apply to “A GPU cluster logs corrected ECC errors increasing steadily on one accelerator although training jobs usually pass”. The controlling explanation is that Rising corrected memory errors may indicate deterioration before uncorrectable failures appear. This question instead requires “PCIe bus errors can disrupt host-to-device transfers even when GPU health itself is normal”.

The related but distinct situation is “A training node reports intermittent GPU Xid faults during collective operations but its nvidia-smi inventory remains normal”. The operational evidence follows from the mechanism that A GPU can enumerate normally while intermittent compute or memory faults occur under particular workloads. For the stated problem, “PCIe bus errors can disrupt host-to-device transfers even when GPU health itself is normal” addresses the central condition.

This alternative addresses the separate case “A multi-GPU server repeatedly loses one NIC link when its fan tray slows under load”. The controlling explanation is that Cooling failures can indirectly destabilize network or accelerator components exposed to excess heat. The deciding requirement here is “PCIe bus errors can disrupt host-to-device transfers even when GPU health itself is normal”.

This statement could explain another scenario: Two otherwise identical optical links use transceiver firmware revisions that negotiate different advertised capabilities on the same switch. In the present case the more specific mechanism is that Part number matching does not by itself prove current operational firmware compatibility.

This option serves the scenario “A fabric upgrade updates all switch operating systems but leaves pluggable optical modules with known compatibility defects”. This case illustrates that Port software and optics form a combined physical signaling system that should be qualified together. This question instead requires “Part number matching does not by itself prove current operational firmware compatibility”.

The option would apply to “An engineer replaces a failed optical module and assumes no post-replacement tests are necessary because the link LED is green”. The key distinction is that A green carrier LED is weaker evidence than validated signal quality at target speed. The present scenario is distinguished by “Part number matching does not by itself prove current operational firmware compatibility”.

This option serves the scenario “A DGX-class server enumerates eight GPUs but one NVLink peer path is missing from the reported topology”. The controlling explanation is that PCIe device discovery does not establish all intended GPU-to-GPU high-bandwidth connections. The present scenario is distinguished by “Logical GPU partitions have specific communication constraints and should not be treated as full physical GPUs by default”.

The option would apply to “A single-node NCCL test works with two GPUs but times out when all eight GPUs participate”. The controlling explanation is that Subset success does not validate the complete intra-node accelerator communication graph. This question instead requires “Logical GPU partitions have specific communication constraints and should not be treated as full physical GPUs by default”.

The related but distinct situation is “NCCL single-node all-reduce bandwidth falls after maintenance even though each GPU still passes a separate CUDA sample”. The controlling explanation is that Collective performance depends on inter-GPU communication in addition to isolated compute health. For the stated problem, “Logical GPU partitions have specific communication constraints and should not be treated as full physical GPUs by default” addresses the central condition.

This alternative addresses the separate case “An operator averages all ClusterKit results into one green score even though one test suite failed every storage path in a rack”. The decisive evidence is that The final readiness report shows independent passes for compute, network and storage checks. For the stated problem, “The final report lists every expected node with a pass, fail or explicit approved exception” addresses the central condition.

The related but distinct situation is “A ClusterKit report shows two hosts consistently below the rack median in GPU-to-network performance”. A completed remediation must demonstrate that After repair the affected nodes fall within approved cluster performance variance. The deciding requirement here is “The final report lists every expected node with a pass, fail or explicit approved exception”.

That observation does not establish operating effectiveness of Reconcile benchmark participants against the complete intended cluster node inventory. The needed evidence is that The final report lists every expected node with a pass, fail or explicit approved exception.

The option would apply to “A high-speed NIC link shows error spikes only after adjacent power cables are rerouted tightly against the data assembly”. The decisive evidence is that The rerouted cable passes the sustained link-quality test with stable error counters. This question instead requires “A cable correction stabilizes pre-FEC and post-FEC indicators during the same sustained test”.

That observation does not establish operating effectiveness of Inspect FEC and symbol counters, optical power and cable condition rather than using carrier state alone. The needed evidence is that A cable correction stabilizes pre-FEC and post-FEC indicators during the same sustained test.

This is weaker evidence because Link-up status can hide marginal high-speed electrical or optical signal integrity. The independent test must show that A cable correction stabilizes pre-FEC and post-FEC indicators during the same sustained test.