NCP-AII Question 90
Select 3You have just completed the installation of NVIDIA GPUs in a server for an AI workload. Which of the following steps should you perform to validate the hardware installation?
- A
Verify GPU presence using the NVIDIA System Management Interface (nvidia-smi).
- B
Run a stress test using a benchmarking tool like NVIDIA DCGM Diagnostics.
- C
Check system BIOS to ensure the GPUs are enabled.
- D
Install the CUDA Toolkit to confirm compatibility with the GPUs.
- E
Verify physical GPU connections by reseating all cables and power connectors.
Show answer and explanation
Correct answers: A, B, C
Explanation
Validating hardware installation involves ensuring the GPUs are recognized by the system, are functional under load, and are enabled in the system BIOS. These steps confirm the hardware is correctly installed and operational. Other tasks, such as installing the CUDA Toolkit or reseating cables, are either post-installation setup or troubleshooting actions, not validation steps.
- A. Correct.
Using the NVIDIA System Management Interface (nvidia-smi) is a critical step to confirm that the system recognizes the installed GPUs and that they are operational.
- B. Correct.
Running a stress test using tools like NVIDIA DCGM Diagnostics helps ensure the GPUs are functioning correctly under load and can handle operational workloads.
- C. Correct.
Checking the system BIOS ensures that the GPUs are enabled and properly configured at the hardware level, which is a necessary validation step.
- D. Incorrect.
While installing the CUDA Toolkit is important for running CUDA-based applications, it is not directly part of validating the hardware installation itself.
- E. Incorrect.
Verifying physical connections is part of the installation process, but it is not a validation step after installation unless issues are identified.