Compute Resources

CamberCloud provides flexible compute resources from small analysis tasks to large-scale GPU simulations. Nova automatically suggests appropriate resources, or you can specify exactly what you need.

Node Sizes

GPU Options

Add GPU acceleration for machine learning and simulations:

@nova Run my training script on a SMALL node with GPU

Available GPU: NVIDIA L4 (24GB VRAM)

Choosing Resources

Let Nova Decide

Simply describe your task and Nova recommends appropriate resources:

@nova I need to process 10GB of genomics data

Specify Explicitly

For precise control:

@nova Run gromacs simulation using MEDIUM node with 4 parallel nodes

Via SDK

job = camber.mpi.create_job(
    command="python train.py",
    node_size="SMALL",
    num_nodes=2,
    with_gpu=True
)

Cost Optimization

• Start small and scale up if needed
• Use spot instances for fault-tolerant workloads
• Monitor job resource usage in logs

Related: Job Attributes | Nodes & Pricing