最新的NVIDIA NCP-AII免費考試真題

問題1

A financial services firm is deploying an AI model for fraud detection that requires rapid inference and data retrieval across multiple sites. Which feature should their storage system prioritize?

A. Low-cost HDD solutions.

B. Tape backup systems.

C. Multi-protocol data access with low latency.

D. High capacity with moderate speed.

正確答案: C

說明：（僅 Fast2test 成員可見）

問題2

An engineer must ensure a BlueField-3 NIC firmware download matches the cluster's PSID.
Which step is critical before installation?

A. Verify the SHA256 hash of the firmware matches NVIDIA's public ledger.

B. Check that the DPU's BMC IP is reachable via ping.

C. Use mstflint -d <PCI_ID> query to validate the firmware's PSID

D. Confirm the firmware file size matches the DPU's flash capacity.

正確答案: C

說明：（僅 Fast2test 成員可見）

問題3

You are tasked with updating both NVIDIA GPU drivers and DOCA drivers on a set of servers used for AI workloads. The environment previously had an older driver stack and custom kernel modules. What is the most important step to successfully upgrade the drivers without causing conflicts?

A. Uninstall all existing GPU and DOCA-related drivers and associated kernel modules before the new install

B. Update the GPU driver leaving the DOCA and OFED drivers unchanged as long as they are detecting the hardware properly.

C. Keep the older driver running alongside the new version in case you need to roll back the upgrade.

D. Validate the driver version post-install since the fresh install will overwrite the legacy drivers.

正確答案: A

說明：（僅 Fast2test 成員可見）

問題4

You are validating the networking infrastructure of a distributed GPU cluster used for AI training with InfiniBand networking fabric installed across nodes. How should you validate node connectivity?

A. Use tools like ib_read_bw and ib_write_bw to test InfiniBand latency and throughput between nodes.

B. Verify IP address assignments across all nodes in the cluster to ensure proper communication paths are established.

C. Confirm that all network switches power on without errors during initialization.

D. Test network performance by transferring files between nodes over Ethernet while monitoring transfer speeds.

正確答案: A

說明：（僅 Fast2test 成員可見）

問題5

A team is validating a DGX BasePOD deployment. Using cmsh, they run a command to check GPU health across all nodes. What indicates that the system is ready for AI workloads?

A. Only the head node's GPUs need to be healthy.

B. All GPUs report Status_Health = OK and Health = OK for each device.

C. At least half of the GPUs report Status_Health = OK

D. The command output is ignored if the system powers on without errors.

正確答案: B

說明：（僅 Fast2test 成員可見）

問題6

The system administrator is tasked with updating GPU drivers and needs to check the driver version that is currently running. Which command should the system administrator use?

A. nvidia-smi --query-gpu=driver_version --format=csv

B. nvidia-smi get gpu driver version --format=csv

C. nvidia-smi gpu-driver-version --format=csv

D. nvidia-smi --query gpu driver --format=csv

正確答案: A

說明：（僅 Fast2test 成員可見）

問題7

An AI operations team enables Multi-Instance GPU (MIG) on several NVIDIA GPUs to improve resource utilization. Shortly afterward, one data science team reports that its distributed training job runs noticeably slower than before. The cluster administrator must determine the most likely reason for this behavior.

A. MIG divides one physical GPU into isolated compute instances with dedicated resources.

B. MIG increases GPU memory bandwidth for every workload.

C. MIG combines multiple physical GPUs into one logical accelerator.

D. MIG automatically enables NVLink aggregation across all GPU instances.

正確答案: A

說明：（僅 Fast2test 成員可見）

問題8

A system administrator is installing a DGX OS and needs to preserve existing configurations and data during installation. What option should be chosen?

A. Export

B. Backup

C. Upgrade

D. Reimage

正確答案: C

說明：（僅 Fast2test 成員可見）

問題9

A cluster administrator deploys GPUDirect RDMA for a distributed AI training environment. After deployment, GPU utilization improves while CPU utilization during communication decreases substantially. Which architectural change most directly explains this improvement?

A. Network traffic is compressed before entering GPU memory.

B. PCIe transactions are replaced entirely by Ethernet broadcasts.

C. All communication is redirected through NVSwitch.

D. Data transfers bypass CPU memory and move directly between GPUs and network adapters.

正確答案: D

說明：（僅 Fast2test 成員可見）

問題10

An administrator installs NVIDIA GPU drivers on a DGX H100 system with UEFI Secure Boot enabled. After reboot, the drivers fail to load. What is the first action to resolve this issue?

A. Reinstall drivers using apt-get install nvidia-driver-550 without rebooting.

B. Disable Secure Boot permanently in BIOS/UEFI settings.

C. Delete /etc/X11/xorg.conf to force driver reconfiguration.

D. Enroll the Machine Owner Key (MOK) during system reboot and enter the recorded password.

正確答案: D

說明：（僅 Fast2test 成員可見）

問題11

For an NVIDIA Enterprise Al Factory with 256 GPUs, which storage solution characteristic is most critical to validate during scaling tests?

A. Consistent per-node throughput ≥8 GiB/s.

B. RAID rebuild times under disk failure.

C. Maximum 4K random read IOPS exceeding 1 million.

D. Single-node write performance during idle clusters.

正確答案: A

說明：（僅 Fast2test 成員可見）

問題12

You are training a deep neural network using NCCL to coordinate communication across four GPUs in a single node. During early performance testing, you notice inconsistent scaling and longer-than-expected training times, even though all GPUs are being used. Which strategy would most effectively improve NCCL efficiency and collective operation performance in this setting?

A. Adjust the batch size so that each GPU receives an equalized portion of the batch, ensuring all GPUs process similar workloads end communication is evenly distributed.

B. Increase the communication frequency between GPUS allowing workloads to be unevenly split, so that synchronization is more frequent and model updates happen faster.

C. Assign the largest possible workload to the first GPU to maximize its utilization, and allow the remaining GPUs to process smaller or variable batch sizes as needed.

D. Disable automatic load balancing so that the deep learning framework can dynamically assign samples to any GPU available during each iteration.

正確答案: A

說明：（僅 Fast2test 成員可見）

問題13

During the setup of a new GPU cluster, you want to validate the GPUs to ensure they meet performance and reliability standards. What would demonstrate performance standards are met?

A. Test a small reference task, such as training a simple model, to confirm general functionality.

B. Validate NVLink bandwidth between GPUs by running data transfer tests across all connected nodes.

C. Verify that the GPU type and count match expected specifications.

D. Use DCGM diagnostics and gpu-burn to stress test GPUs under load.

正確答案: D

說明：（僅 Fast2test 成員可見）

問題14

On a DCX, a system administrator needs to monitor PSU, CPU Utilization, GPU Utilization, RAID, and Memory Utilization. What single NVIDIA tool should be used?

A. NVSM

B. MOFED

C. DCGM

D. SMI

正確答案: A

說明：（僅 Fast2test 成員可見）

問題15

A user encounters "permission denied" errors when running GPU-accelerated containers on a Secure Boot-enabled system. What resolves this?

A. Reinstall Docker without the NVIDIA runtime.

B. Run Docker with sudofor elevated privileges.

C. Enroll the MOK and sign NVIDIA kernel modules.

D. Disable SELinux to relax unnecessary security policies.

正確答案: C

說明：（僅 Fast2test 成員可見）

最新的NVIDIA AI Infrastructure - NCP-AII免費考試真題

聯系我們

站內鏈接

最新更新