Before installing RAPIDS, an engineer must confirm that the workstation's NVIDIA driver supports the CUDA version that the chosen RAPIDS build needs. Which command reports the installed driver along with the maximum CUDA version that driver can support?
- Apip list, because it prints the CUDA runtime libraries bundled inside the active Python environment along with the host driver version.
- Bconda info, because it summarises the conda installation and reports the GPU driver detected on the host machine.
- Cnvidia-smi, because it queries the installed NVIDIA driver and reports the driver version along with the maximum CUDA version that driver supports. Correct
- Dpython --version, because the interpreter version determines which CUDA-enabled RAPIDS builds the host driver will accept.
Why A is wrong: pip list shows installed Python packages in the environment, which can include CUDA wheels, but it reports nothing about the system NVIDIA driver or the CUDA version that driver supports.
Why B is wrong: conda info describes the conda configuration and channels, which is useful context, but it does not query the NVIDIA driver or report the driver's supported CUDA version.
Why C is correct: nvidia-smi talks to the installed driver and prints the driver version and the highest CUDA version it can support, which is exactly the compatibility check needed before selecting a RAPIDS build.
Why D is wrong: The Python version constrains which package builds you can install, but it says nothing about the GPU driver or its supported CUDA version, so it cannot perform the driver compatibility check.