Interactive Session (srun):
# For A100 GPUs:
srun --gres=gpu:a100:1 --pty bash # Request 1 A100 GPU
srun --gres=gpu:a100:4 --pty bash # Request 4 A100 GPUs
# For A30 GPUs:
srun --gres=gpu:a30:1 --pty bash # Request 1 A30 GPU
srun --gres=gpu:a30:4 --pty bash # Request 4 A30 GPUsBatch Jobs (sbatch):
#!/bin/bash
#SBATCH --job-name=gpu_job
#SBATCH --output=output_%j.log
#SBATCH --error=error_%j.log
#SBATCH --time=1:00:00 # HH:MM:SS
#SBATCH --gres=gpu:a100:2 # Request 2 A100 GPUs
# Your commands here
nvidia-smi # Check GPU allocation
python your_script.py--gres=gpu:a100:1 or --gres=gpu:a30:1 for single GPU tasksnvidia-smi # Show GPU usage
nvidia-smi -l 1 # Update every second