Online Access Free NCP-AII Practice Test

You are managing a cluster of servers running Docker and NVIDIA GPUs. You want to monitor the GPU utilization of all Docker containers running on the cluster in real-time. Which tools or techniques could you use to achieve this?

• A.Integrate the DCGM exporter with Prometheus and Grafana to visualize GPU metrics from all nodes in the cluster.
• B.Use 'nvidia-smi' on each host and parse the output to extract GPIJ utilization for each container.
• C.Utilize NVIDIA's NGC cloud monitoring services to automatically collect and visualize GPU metrics from all nodes.
• D.Implement a custom script that uses the NVIDIA Management Library (NVML) to query GPU utilization from each container.
• E.Use 'docker statS command on each host and look for the GPIJ utilization metrics.

You are designing a storage system using BeeGFS for an AI cluster. The cluster consists of 10 client nodes, each with 2 NVIDIAAIOO GPUs, and 4 storage servers. Each storage server has 10 NVMe SSDs. The training dataset is 100TB. You want to ensure high availability and performance. Which of the following BeeGFS configurations would be MOST appropriate?

• A.10 MDS server (one per client nodes) and a single OST.
• B.Four MDS servers (one per storage server) and 40 OSTs (one per NVMe SSD).
• C.A single metadata server (MDS) and four storage targets (OSTs), each spanning all 10 NVMe SSDs on a storage server.
• D.One MDS server and 10 OSTs, splitting each NVMe SSD into smaller virtual OSTs.
• E.Two MDS servers in a high-availability configuration and 40 OSTs (one per NVMe SSD).

Which command is used to enable MIG mode on an NVIDIAAmpere architecture GPU?

• A.nvidia-smi -set-mig-mode=enable -i 0
• B.nvidia-smi -gpu 0 -i MIG=I
• C.nvidia-smi -i 0 -mig 1
• D.nvidia-smi -i 0 -enable-mig
• E.nvidia-smi -format=csv,noheader=l -i 0 -enable-mig

An AI inference workload running on a server equipped with NVIDIA TensorRT is experiencing intermittent performance degradation. Profiling reveals potential issues with GPU utilization. Which command-line tool is best suited to provide real-time GPU performance metrics, including utilization, memory usage, and power consumption, during workload execution?

• A.tegrastats
• B.nvitop
• C.nvprof
• D.gpustat
• E.nvidia-smi

You are deploying an NVIDIA GPU-accelerated application in a virtualized environment using vGPU. How does vGPU technology impact power and cooling considerations compared to a bare-metal deployment, and what specific monitoring metrics become crucial?

• A.vGPU deployments eliminate the need for GPU monitoring, as the virtualization layer handles all power and cooling management.
• B.vGPU deployments have no significant impact on power and cooling requirements compared to bare-metal. Standard GPU temperature and power draw metrics are sufficient.
• C.vGPU deployments can lead to higher overall power consumption and concentrated heat generation on the host server due to resource consolidation. Monitoring metrics like GPU utilization per VM, vGPU frame rate, and host server thermal headroom become crucial.
• D.vGPU deployments typically require less power and cooling than bare-metal, as resources are shared. The host server's overall power consumption becomes the primary monitoring metric.
• E.vGPU deployments require specialized cooling solutions that are not needed for bare metal setups.