Skip to main content
GPU Cloud provides dedicated compute infrastructure for machine learning workloads. Use GPU clusters to train models, run inference, and process large-scale AI tasks.

What is a GPU cluster

A GPU cluster is a group of interconnected servers, each equipped with multiple high-performance GPUs. Clusters are designed for workloads that require massive parallel processing power, such as training large language models (LLMs), fine-tuning foundation models, running inference at scale, and high-performance computing (HPC) tasks.
GPU Cloud create cluster page showing region selection, cluster type, and GPU configuration options
All nodes in a cluster share the same configuration: operating system image, network settings, and storage mounts. This ensures consistent behavior across the cluster.

Cluster types

Gcore offers three types of GPU clusters:
TypeDescriptionBest for
Bare Metal GPUDedicated physical servers with guaranteed resources. No virtualization overheadProduction workloads, long-running training jobs, and latency-sensitive inference
Spot Bare Metal GPUSame hardware as Bare Metal, but at a reduced price (up to 50% discount). Instances can be preempted with a 24-hour notice when capacity is neededFault-tolerant training with checkpointing, batch processing, development, and testing
Virtual GPUVirtualized GPU instances with flexible resource management. Supports flavor changes and cost optimization through shelving (powering off releases resources and stops billing)Development environments, variable workloads, cost-sensitive projects
Clusters can scale to hundreds of nodes. Production deployments with 250+ nodes in a single cluster are supported, limited only by regional stock availability.

Available configurations

Select a configuration based on your workload requirements:
ConfigurationGPUsInterconnectRAMStorageUse case
H200 with InfiniBand8x NVIDIA H200 141GB3.2 Tbit/s InfiniBand, 2x 200 Gbit/s Ethernet2TB6x 3.84TB NVMeDistributed LLM training with the latest GPU generation
H100 with InfiniBand8x NVIDIA H100 80GB3.2 Tbit/s InfiniBand2TBup to 8x 3.84TB NVMeDistributed LLM training requiring high-speed inter-node communication
A100 with InfiniBand8x NVIDIA A100 80GB800 Gbit/s InfiniBand2TB8x 3.84TB NVMeMulti-node ML training and HPC workloads
A100 without InfiniBand8x NVIDIA A100 80GB2x 100 Gbit/s Ethernet2TB8x 3.84TB NVMeSingle-node training, inference for large models requiring more than 48GB VRAM
L40S8x NVIDIA L40S2x 25 Gbit/s Ethernet2TB4x 7.68TB NVMeInference, fine-tuning small to medium models requiring less than 48GB VRAM
Outbound data transfer (egress) from GPU clusters is free. Other costs are covered in GPU Cloud billing.

InfiniBand networking

InfiniBand is a high-bandwidth, low-latency interconnect technology used for communication between nodes in a cluster. InfiniBand is configured automatically when you create a cluster. If the selected configuration includes InfiniBand network cards, all nodes are placed in the same InfiniBand domain with no manual setup required. H100 configurations typically have 8 InfiniBand ports per node, each creating a dedicated network interface. Gcore includes SHARP (Scalable Hierarchical Aggregation and Reduction Protocol) automatically for InfiniBand configurations. It offloads collective communication operations to network switches, reducing latency for HPC and AI workloads. InfiniBand matters most for distributed training, where models that don’t fit on a single node require frequent gradient synchronization between GPUs. The same applies to multi-node inference when large models are split across servers. In these cases, InfiniBand reduces communication overhead significantly compared to Ethernet. For single-node workloads or independent batch jobs that don’t require node-to-node communication, InfiniBand provides no benefit. Standard Ethernet configurations work equally well and may be more cost-effective.

Storage options

GPU clusters support two storage types:
Storage typePersistencePerformanceUse case
Local NVMeTemporary (deleted with cluster)Highest IOPS, lowest latencyTraining data cache, checkpoints during training
File sharesPersistent (independent of cluster)Network-attached, lower latency than object storageDatasets, model weights, shared checkpoints

Cluster lifecycle

Create --> Configure --> Run workloads --> Resize (optional) --> Delete
  1. Create: Select region, GPU type, number of nodes, image, and network settings. See creating a Bare Metal GPU cluster or creating a Virtual GPU cluster.
  2. Configure: Connect via SSH to each node, install required dependencies, and mount file shares to prepare the environment for workloads.
  3. Run workloads: Execute training jobs, run inference services, process data.
  4. Resize: Add or remove nodes based on demand. New nodes inherit the cluster configuration. See managing a Bare Metal GPU cluster for details.
  5. Delete: Remove the cluster when no longer needed. Local storage is erased; file shares and network disks can be preserved.

GPU cluster characteristics

  • Provisioning takes 15–40 minutes
  • You define the configuration (image, network, and storage) at creation, and you cannot change it afterward
  • Local NVMe storage is temporary—store critical data in persistent file shares
  • Spot clusters can be interrupted with a 24-hour notice
  • Available regional capacity determines cluster size
  • Servers equipped with BlueField network cards support hardware firewalls

FAQ

When should I use a GPU cluster?

Use a cluster when your workload requires more resources than a single node can provide or when scaling across multiple machines for distributed training or inference is required.

What is the difference between a single GPU server and a cluster?

A single GPU server runs workloads on one node. A cluster combines multiple nodes into a single environment:
  • Single GPU server: 1–8 GPUs on one node
  • Cluster: Multiple nodes with 16+ GPUs in total
Use a single GPU server for development or smaller workloads. Use a cluster for large models, distributed training, or high-scale inference.

How do I verify that my cluster is working correctly?

Connect to a node using SSH and run:
  • nvidia-smi — confirm GPU availability
  • ibstat — verify InfiniBand connectivity
If your cluster uses InfiniBand, nodes connect within the same network domain.

How does a floating IP work with a GPU cluster?

A floating IP points to a single node in the cluster. Use that node as an entry point:
  1. Connect via SSH using the floating IP
  2. Access other nodes through their internal IP addresses
A floating IP does not represent the entire cluster.

Does GPU Cloud work as a serverless platform?

GPU Cloud provides dedicated infrastructure. When you create a cluster:
  • You provision fixed GPU and compute resources
  • The cluster runs continuously until you resize or delete it