/gpu-status

Check GPU passthrough and NVIDIA GPU status:

1. Verify GPU is visible: lspci | grep -i nvidia
2. Check NVIDIA driver status: nvidia-smi
3. Review GPU utilization: nvidia-smi --query-gpu=utilization.gpu,utilization.memory,temperature.gpu --format=csv
4. Check GPU processes: nvidia-smi pmon
5. Review NVIDIA driver version: nvidia-smi --query-gpu=driver_version --format=csv,noheader
6. Check for GPU errors: dmesg | grep -i nvidia
7. Verify CUDA availability if needed: nvcc --version 2>/dev/null || echo "CUDA not installed"
8. Check GPU memory: nvidia-smi --query-gpu=memory.used,memory.total --format=csv
9. Review GPU configuration in VM: lspci -v -s $(lspci | grep -i nvidia | cut -d' ' -f1)

Provide Daniel with:

• GPU model and driver version
• GPU utilization (compute and memory)
• Temperature status
• Running processes using GPU
• GPU memory usage
• Any driver errors or warnings
• CUDA availability and version
• Passthrough configuration status
• Recommendations for GPU optimization
• Warnings if GPU is overheating or overutilized