NCA - AI Infrastructure and Operations NCA-AIIO Practice Test 60
Question 590
A GPU operations plan calls for: Explore unsupervised clustering using selected normalized features and verify cluster interpretability. The reason is that an operations team has telemetry for thousands of devices but no reliable failure labels and wants to group similar behavior. Which principle is MOST relevant?
- Unsupervised learning discovers structure without relying on a target label for every example.
- Deep learning typically involves neural networks with multiple learned representation layers and high compute demand.
- Machine learning uses data-driven parameter fitting rather than solely hand-authored conditional decision rules.
- Inference normally applies learned parameters; serving requests alone does not imply those parameters update.
Explanation
Correct Answer
A:The governing principle is that unsupervised learning discovers structure without relying on a target label for every example. It supports the measure to explore unsupervised clustering using selected normalized features and verify cluster interpretability. because an operations team has telemetry for thousands of devices but no reliable failure labels and wants to group similar behavior. Without the control, the team may invent labels or treat arbitrary clusters as proven hardware-failure predictions. The most useful acceptance evidence is: A sample review establishes meaningful cluster properties without claiming the clusters predict confirmed incidents.
Incorrect Answers
B:That principle is relevant to this distinct case: An enterprise compares a decision tree with a deep multilayer vision network and claims both use identical learning architectures. The associated response would be to classify deep learning as a subset of machine learning using layered neural representations. It does not describe the central reason for this case, which is that unsupervised learning discovers structure without relying on a target label for every example.
C:That principle is relevant to this distinct case: A project hard-codes hundreds of manually chosen rules to classify support tickets, yet the team describes it as a trained ML model. The associated response would be to distinguish fixed rule-based automation from models that learn patterns from examples during training. It does not describe the central reason for this case, which is that unsupervised learning discovers structure without relying on a target label for every example.
D:That principle is relevant to this distinct case: A company says its production fraud-scoring API is still learning from each request although model weights never change. The associated response would be to separate trained-model inference from explicit retraining or online-learning update processes. It does not describe the central reason for this case, which is that unsupervised learning discovers structure without relying on a target label for every example.
Question 591
The proposed approach is to distinguish fixed rule-based automation from models that learn patterns from examples during training. In the following setting, a project hard-codes hundreds of manually chosen rules to classify support tickets, yet the team describes it as a trained ML model. Which principle BEST supports the approach?
- Unsupervised learning discovers structure without relying on a target label for every example.
- Reinforcement learning improves an action policy using feedback from an environment rather than only labeled examples.
- Supervised learning learns to map inputs to known target labels rather than discover unlabeled groups.
- Machine learning uses data-driven parameter fitting rather than solely hand-authored conditional decision rules.
Explanation
Correct Answer
D:The governing principle is that machine learning uses data-driven parameter fitting rather than solely hand-authored conditional decision rules. It supports the measure to distinguish fixed rule-based automation from models that learn patterns from examples during training. because a project hard-codes hundreds of manually chosen rules to classify support tickets, yet the team describes it as a trained ML model. Without the control, the project may promise improvements from additional data even though its rule engine never learns. The most useful acceptance evidence is: A review identifies the rule configuration and confirms there is no data-driven parameter optimization step.
Incorrect Answers
A:That principle is relevant to this distinct case: An operations team has telemetry for thousands of devices but no reliable failure labels and wants to group similar behavior. The associated response would be to explore unsupervised clustering using selected normalized features and verify cluster interpretability. It does not describe the central reason for this case, which is that machine learning uses data-driven parameter fitting rather than solely hand-authored conditional decision rules.
B:That principle is relevant to this distinct case: A robot learns control actions by interacting with a simulated environment and receiving delayed reward signals. The associated response would be to recognize reinforcement learning and define rewards, actions and safe simulation boundaries. It does not describe the central reason for this case, which is that machine learning uses data-driven parameter fitting rather than solely hand-authored conditional decision rules.
C:That principle is relevant to this distinct case: A bank has historical transactions labeled as legitimate or fraudulent and wants to train a classifier from those examples. The associated response would be to choose a supervised-learning workflow with labeled train, validation and test partitions. It does not describe the central reason for this case, which is that machine learning uses data-driven parameter fitting rather than solely hand-authored conditional decision rules.
Question 592
The selected AI infrastructure safeguard is not implemented: Validate truly independent power paths and loss-of-feed operation with qualified facility engineers. The observed situation is: A cluster uses redundant electrical feeds but both power cables are plugged into the same unprotected PDU. What exposure remains?
- Loss of cooling under peak GPU load can produce widespread thermal alarms and abrupt job termination.
- Hotspots can cause thermal throttling or emergency shutdown while overall room temperature seems normal.
- One PDU failure may still switch off both power supplies and stop the cluster.
- One chassis can throttle or shut down without a building-wide temperature alarm.
Explanation
Correct Answer
C:The primary exposure is that one PDU failure may still switch off both power supplies and stop the cluster. This follows from a cluster uses redundant electrical feeds but both power cables are plugged into the same unprotected PDU. where the omitted measure was to validate truly independent power paths and loss-of-feed operation with qualified facility engineers. The technical reason is that redundancy only mitigates a failure if the supposedly independent paths avoid the same fault domain. A good operational validation after repair would be: A controlled failover plan demonstrates supported equipment remains powered after either approved feed is lost.
Incorrect Answers
A:This exposure arises in a different situation: Facilities schedules maintenance to an air-handling unit while operations keeps every GPU rack at full demand. There, the corresponding safeguard is to coordinate workload reduction and protected shutdown windows against available cooling redundancy. In the present case, the more direct risk is that one PDU failure may still switch off both power supplies and stop the cluster.
B:This exposure arises in a different situation: An existing server room can cool ordinary web racks but a planned accelerated rack doubles local heat density. There, the corresponding safeguard is to validate rack heat rejection and cooling solution against documented maximum system power. In the present case, the more direct risk is that one PDU failure may still switch off both power supplies and stop the cluster.
D:This exposure arises in a different situation: A rack runs multiple powerful accelerators but operators monitor only average room temperature, not intake air. There, the corresponding safeguard is to collect rack inlet and component thermal measurements during representative peak workloads. In the present case, the more direct risk is that one PDU failure may still switch off both power supplies and stop the cluster.
Question 593
The proposed approach is to size online inference for bounded tail latency, concurrent sessions and availability at the actual model size. In the following setting, a hospital uses a deployed vision model to classify imaging requests within a fixed response time during business hours. Which principle BEST supports the approach?
- Inference runs a trained model for predictions and is often governed by response-time service-level objectives.
- Training repeatedly computes gradients and updates parameters, placing different demands on hardware than inference.
- Inference scale-out must include model loading and initialization time, not only GPU allocation time.
- Batching can improve throughput but adds queueing latency and must respect serving requirements.
Explanation
Correct Answer
A:The governing principle is that inference runs a trained model for predictions and is often governed by response-time service-level objectives. It supports the measure to size online inference for bounded tail latency, concurrent sessions and availability at the actual model size. because a hospital uses a deployed vision model to classify imaging requests within a fixed response time during business hours. Without the control, a deployment optimized only for bulk throughput can violate urgent per-request latency requirements. The most useful acceptance evidence is: A simulated concurrent request load stays below its p95 deadline and preserves agreed model output accuracy.
Incorrect Answers
B:That principle is relevant to this distinct case: A research team trains a large image classifier nightly with lengthy epochs and no strict per-request response deadline. The associated response would be to optimize for sustained training throughput, memory capacity and batch processing rather than single-request latency. It does not describe the central reason for this case, which is that inference runs a trained model for predictions and is often governed by response-time service-level objectives.
C:That principle is relevant to this distinct case: An autoscaled text-generation endpoint starts new model replicas only after a demand spike and sees long cold starts. The associated response would be to preload supported model artifacts or maintain warm capacity based on startup latency and expected demand. It does not describe the central reason for this case, which is that inference runs a trained model for predictions and is often governed by response-time service-level objectives.
D:That principle is relevant to this distinct case: An inference service receives many short requests but operates each GPU one request at a time with poor utilization. The associated response would be to evaluate dynamic batching within the permitted queueing delay and model-specific latency budget. It does not describe the central reason for this case, which is that inference runs a trained model for predictions and is often governed by response-time service-level objectives.
Question 594
The team recommends this safeguard: Recognize the role of widely available frameworks and reusable libraries in expanding AI adoption. It applies because two groups can now build the same neural architecture using mature open frameworks instead of custom kernels. Which concept justifies the design?
- Accessible high-level frameworks reduce implementation barriers around model training and deployment.
- On-demand infrastructure reduces initial capital barriers to experimentation for intermittent workloads.
- Recent AI improvements depend on useful training data as well as computing resources and algorithms.
- Reusable pretrained representations reduce the data and compute needed for many specialized tasks.
Explanation
Correct Answer
A:The governing principle is that accessible high-level frameworks reduce implementation barriers around model training and deployment. It supports the measure to recognize the role of widely available frameworks and reusable libraries in expanding AI adoption. because two groups can now build the same neural architecture using mature open frameworks instead of custom kernels. Without the control, teams may spend capacity replicating common primitives rather than validating their actual use cases. The most useful acceptance evidence is: A supported framework implementation reproduces a documented reference baseline with recorded versions.
Incorrect Answers
B:That principle is relevant to this distinct case: A startup needs accelerator capacity only during monthly training bursts and cannot finance an idle cluster. The associated response would be to evaluate elastic cloud accelerator access as an adoption enabler while recording ongoing costs and data transfer. It does not describe the central reason for this case, which is that accessible high-level frameworks reduce implementation barriers around model training and deployment.
C:That principle is relevant to this distinct case: A recommendation project improves after collecting representative interaction logs but management attributes all gains to GPU upgrades. The associated response would be to evaluate data quantity, label quality and dataset representativeness alongside accelerated hardware improvements. It does not describe the central reason for this case, which is that accessible high-level frameworks reduce implementation barriers around model training and deployment.
D:That principle is relevant to this distinct case: An applied AI team starts from an existing pretrained vision model rather than training its representation from random weights. The associated response would be to use transfer learning appropriately and validate downstream quality with the domain-specific dataset. It does not describe the central reason for this case, which is that accessible high-level frameworks reduce implementation barriers around model training and deployment.
Question 595
The selected AI infrastructure safeguard is not implemented: Use an appropriate workload scheduler with GPU resource accounting and topology-aware placement. The observed situation is: A shared accelerator fleet has idle GPUs in some nodes while large training jobs wait without proper scheduling. What exposure remains?
- The new fleet may deliver little additional model throughput while collective traffic saturates links.
- Expensive devices can sit idle while requests are stranded behind bad resource placement.
- Distributed jobs may hit external network limits even when local GPU links have ample bandwidth.
- A single failed node can postpone critical jobs for days when the cluster runs fully booked.
Explanation
Correct Answer
B:The primary exposure is that expensive devices can sit idle while requests are stranded behind bad resource placement. This follows from a shared accelerator fleet has idle GPUs in some nodes while large training jobs wait without proper scheduling. where the omitted measure was to use an appropriate workload scheduler with GPU resource accounting and topology-aware placement. The technical reason is that gPU cluster scaling requires coordinated allocation of devices and jobs, not just adding hardware. A good operational validation after repair would be: A representative mix of jobs shows better allocated GPU utilization and bounded queueing delay.
Incorrect Answers
A:This exposure arises in a different situation: A data center scales from four to sixty-four training nodes without upgrading per-node network ports. There, the corresponding safeguard is to model aggregate east-west traffic and network oversubscription at the intended cluster scale. In the present case, the more direct risk is that expensive devices can sit idle while requests are stranded behind bad resource placement.
C:This exposure arises in a different situation: A team adds more GPUs inside one system and assumes the same link topology extends to GPUs in other racks. There, the corresponding safeguard is to distinguish node-local NVLink or NVSwitch scaling from networked multi-node scale-out design. In the present case, the more direct risk is that expensive devices can sit idle while requests are stranded behind bad resource placement.
D:This exposure arises in a different situation: A research group buys all available GPUs with no spare capacity for maintenance or failed accelerator replacement. There, the corresponding safeguard is to define capacity headroom and controlled drain/replacement procedures based on job availability targets. In the present case, the more direct risk is that expensive devices can sit idle while requests are stranded behind bad resource placement.
Question 596
An operations team implements: Include checkpoint size, frequency, retention and restore throughput in capacity and storage planning. Which test result would MOST convincingly show the approach works?
- A checkpoint-and-restore exercise completes inside recovery targets with sufficient storage headroom.
- An instrumented training step confirms peak optimizer, gradient and activation allocations fit capacity.
- A topology-aware test reports lower communication latency after matching GPU, CPU and NIC locality.
- The data-loader benchmark maintains intended batch supply without swap pressure or CPU queue buildup.
Explanation
Correct Answer
A:The strongest proof is: A checkpoint-and-restore exercise completes inside recovery targets with sufficient storage headroom. It directly tests the safeguard to include checkpoint size, frequency, retention and restore throughput in capacity and storage planning. The control addresses this specific context: A training program uses frequent multi-terabyte checkpoints but reserves storage for only its original dataset. It matters because reliable training infrastructure must store and restore evolving state as well as original training data. Without it, a full storage volume can interrupt runs or prevent recovery after a node failure.
Incorrect Answers
B:This observation would validate a different control: Model total training-state memory and evaluate supported sharding or memory-saving strategies. It fits the case where a team estimates LLM training memory from the file size of weights and omits optimizer and gradient storage. It cannot prove success here; instead the required evidence is a checkpoint-and-restore exercise completes inside recovery targets with sufficient storage headroom.
C:This observation would validate a different control: Inspect CPU NUMA, PCIe and GPU-to-NIC affinity before assigning communication-heavy training jobs. It fits the case where two GPUs share a host but one is connected through a constrained PCIe path to the preferred network adapter. It cannot prove success here; instead the required evidence is a checkpoint-and-restore exercise completes inside recovery targets with sufficient storage headroom.
D:This observation would validate a different control: Size host RAM and CPU data-loader workers for preprocessing, buffering and model-feed requirements. It fits the case where a GPU server repeatedly runs out of host memory when constructing large mini-batches from many source files. It cannot prove success here; instead the required evidence is a checkpoint-and-restore exercise completes inside recovery targets with sufficient storage headroom.
Question 597
The following safeguard was deployed: Coordinate workload reduction and protected shutdown windows against available cooling redundancy. What operational observation would demonstrate its success?
- Measured inlet temperatures and predicted rack heat remain inside approved vendor and facility limits.
- An approved maintenance rehearsal holds measured rack temperatures within safe limits during reduced cooling.
- A facilities-approved electrical load plan meets breaker, feed and redundant-path capacity requirements.
- A heat-load test records compliant inlet, GPU temperature and clock behavior for each high-density node.
Explanation
Correct Answer
B:The strongest proof is: An approved maintenance rehearsal holds measured rack temperatures within safe limits during reduced cooling. It directly tests the safeguard to coordinate workload reduction and protected shutdown windows against available cooling redundancy. The control addresses this specific context: Facilities schedules maintenance to an air-handling unit while operations keeps every GPU rack at full demand. It matters because data-center maintenance planning must account for the coupling between compute load and cooling availability. Without it, loss of cooling under peak GPU load can produce widespread thermal alarms and abrupt job termination.
Incorrect Answers
A:This observation would validate a different control: Validate rack heat rejection and cooling solution against documented maximum system power. It fits the case where an existing server room can cool ordinary web racks but a planned accelerated rack doubles local heat density. It cannot prove success here; instead the required evidence is an approved maintenance rehearsal holds measured rack temperatures within safe limits during reduced cooling.
C:This observation would validate a different control: Calculate worst-case electrical demand and verify PDU, branch-circuit and feed ratings with facilities staff. It fits the case where a proposed rack contains high-density GPU servers but its power distribution units cannot deliver full simultaneous load. It cannot prove success here; instead the required evidence is an approved maintenance rehearsal holds measured rack temperatures within safe limits during reduced cooling.
D:This observation would validate a different control: Collect rack inlet and component thermal measurements during representative peak workloads. It fits the case where a rack runs multiple powerful accelerators but operators monitor only average room temperature, not intake air. It cannot prove success here; instead the required evidence is an approved maintenance rehearsal holds measured rack temperatures within safe limits during reduced cooling.
Question 598
The design team faces this case: An enterprise compares a decision tree with a deep multilayer vision network and claims both use identical learning architectures. It adopts: Classify deep learning as a subset of machine learning using layered neural representations. What evidence BEST validates the intended outcome?
- A simulator run records state transitions, actions and accumulated rewards under the intended policy.
- The design document distinguishes shallow ML features from learned neural representations and records resource implications.
- The evaluation suite scores factual support of summaries separately from the department classification accuracy.
- A sample review establishes meaningful cluster properties without claiming the clusters predict confirmed incidents.
Explanation
Correct Answer
B:The strongest proof is: The design document distinguishes shallow ML features from learned neural representations and records resource implications. It directly tests the safeguard to classify deep learning as a subset of machine learning using layered neural representations. The control addresses this specific context: An enterprise compares a decision tree with a deep multilayer vision network and claims both use identical learning architectures. It matters because deep learning typically involves neural networks with multiple learned representation layers and high compute demand. Without it, capacity planning may ignore the different training memory and accelerator needs of a deep network.
Incorrect Answers
A:This observation would validate a different control: Recognize reinforcement learning and define rewards, actions and safe simulation boundaries. It fits the case where a robot learns control actions by interacting with a simulated environment and receiving delayed reward signals. It cannot prove success here; instead the required evidence is the design document distinguishes shallow ML features from learned neural representations and records resource implications.
C:This observation would validate a different control: Distinguish generative modeling from discriminative classification when selecting model evaluation criteria. It fits the case where a text model generates new summaries while a separate classifier assigns each message to a fixed department. It cannot prove success here; instead the required evidence is the design document distinguishes shallow ML features from learned neural representations and records resource implications.
D:This observation would validate a different control: Explore unsupervised clustering using selected normalized features and verify cluster interpretability. It fits the case where an operations team has telemetry for thousands of devices but no reliable failure labels and wants to group similar behavior. It cannot prove success here; instead the required evidence is the design document distinguishes shallow ML features from learned neural representations and records resource implications.
Question 599
An architecture review proposes: Enforce supported device allocation, tenant permissions and container isolation for GPU workloads. The underlying requirement is: An inference container requests a GPU but can access every accelerator and shared host filesystem on its node. What technical rationale MOST strongly supports it?
- Multi-tenant accelerator scheduling must balance fairness and business priority rather than first-come volume alone.
- Workload lifecycle coordination prevents maintenance actions from racing with new GPU job placement.
- Scheduler resource claims and runtime security controls jointly determine workload isolation.
- Kubernetes requires an appropriate device discovery and allocation path to schedule accelerator workloads.
Explanation
Correct Answer
C:The governing principle is that scheduler resource claims and runtime security controls jointly determine workload isolation. It supports the measure to enforce supported device allocation, tenant permissions and container isolation for GPU workloads. because an inference container requests a GPU but can access every accelerator and shared host filesystem on its node. Without the control, one tenant may access unassigned accelerator capacity or sensitive shared data. The most useful acceptance evidence is: A negative access test confirms the workload sees only approved GPUs and mounted storage.
Incorrect Answers
A:That principle is relevant to this distinct case: Multiple research teams compete for a shared GPU fleet and one account submits hundreds of low-priority jobs. The associated response would be to define resource quotas, scheduling priorities and fair-share policy with accountable ownership. It does not describe the central reason for this case, which is that scheduler resource claims and runtime security controls jointly determine workload isolation.
B:That principle is relevant to this distinct case: A GPU node is marked for maintenance while the scheduler continues assigning new long-running training jobs to it. The associated response would be to cordon or drain the node through approved scheduler workflows and coordinate checkpointable tasks. It does not describe the central reason for this case, which is that scheduler resource claims and runtime security controls jointly determine workload isolation.
D:That principle is relevant to this distinct case: A Kubernetes AI cluster contains NVIDIA GPUs but pod resource requests are invisible to the control plane. The associated response would be to install and validate supported GPU drivers, NVIDIA device integration and resource advertisement. It does not describe the central reason for this case, which is that scheduler resource claims and runtime security controls jointly determine workload isolation.
Question 600
The team recommends this safeguard: Separate trained-model inference from explicit retraining or online-learning update processes. It applies because a company says its production fraud-scoring API is still learning from each request although model weights never change. Which concept justifies the design?
- Deep learning typically involves neural networks with multiple learned representation layers and high compute demand.
- Generative systems create content or sequences while classifiers predict categories for observed inputs.
- Machine learning uses data-driven parameter fitting rather than solely hand-authored conditional decision rules.
- Inference normally applies learned parameters; serving requests alone does not imply those parameters update.
Explanation
Correct Answer
D:The governing principle is that inference normally applies learned parameters; serving requests alone does not imply those parameters update. It supports the measure to separate trained-model inference from explicit retraining or online-learning update processes. because a company says its production fraud-scoring API is still learning from each request although model weights never change. Without the control, operators can overlook model drift while assuming a deployed checkpoint adapts automatically. The most useful acceptance evidence is: A model-version audit shows fixed parameters during inference and a separate governed update process.
Incorrect Answers
A:That principle is relevant to this distinct case: An enterprise compares a decision tree with a deep multilayer vision network and claims both use identical learning architectures. The associated response would be to classify deep learning as a subset of machine learning using layered neural representations. It does not describe the central reason for this case, which is that inference normally applies learned parameters; serving requests alone does not imply those parameters update.
B:That principle is relevant to this distinct case: A text model generates new summaries while a separate classifier assigns each message to a fixed department. The associated response would be to distinguish generative modeling from discriminative classification when selecting model evaluation criteria. It does not describe the central reason for this case, which is that inference normally applies learned parameters; serving requests alone does not imply those parameters update.
C:That principle is relevant to this distinct case: A project hard-codes hundreds of manually chosen rules to classify support tickets, yet the team describes it as a trained ML model. The associated response would be to distinguish fixed rule-based automation from models that learn patterns from examples during training. It does not describe the central reason for this case, which is that inference normally applies learned parameters; serving requests alone does not imply those parameters update.