NVIDIA-Certified Professional: AI Operations 試験対策問題集

模擬試験モードの設定

上記は本サイトの模擬試験モードの設定値です。実際の試験の出題数・制限時間・合格基準は、各認定プログラムの公式サイトでご確認ください。

出題分野

• Installation and Deployment（導入と展開）（出題比率 30%）
• Administration（管理）（出題比率 24%）
• Workload Management（ワークロード管理）（出題比率 23%）
• Troubleshooting and Optimization（障害切り分けと最適化）（出題比率 23%）

フリー問題

1. A team is specifying shared storage for a new training cluster. A vendor propose…
2. An administrator sets a workload manager client parameter on a node category. A …
3. A security fix has been applied inside the software image used by 30 running com…
4. Each compute node has a large local drive. A team asks whether the dataset shoul…
5. A user on a compute node tries to pull a framework container from the vendor reg…
6. A newly racked node netboots but the operator sees it sit in a state whose statu…
7. A cluster runs many single-accelerator jobs, and the operations team finds that …
8. An administrator wants only six of the eight accelerators on a node to be schedu…
9. During an operational review of an accelerated data center, a manager asks which…
10. A developer wants a container in the orchestration cluster to receive one accele…
11. On a freshly configured queue-based cluster, users report that their jobs land o…
12. A platform team must keep three application teams from taking each other's capac…
13. Preemptible services on the resource allocation platform are being stopped and r…
14. A team on the resource allocation platform has an assigned share of eight accele…
15. A researcher submits a batch script that asks for four accelerators on one node.…
16. A user's training script indexes accelerators from zero and works when it gets t…
17. A multi-process data loader inside a training container fails with an error abou…
18. A team confirms the direct storage path works for large reads but finds that sma…
19. A site keeps hitting the failure where containers lose their accelerators after …
20. A site configures a degraded-mode policy so that a link failure does not take th…