GPU Compute

GPU Model	Configuration	Price	Memory	Key Features
H200 NVL PCI-E	4x or 8x GPUs	$2.70/hour	-	Single-node with NVLink
H100 NVL PCI-E	2x, 4x, or 8x GPUs	$1.88/hour	-	Single-node with NVLink
HGX H200	4x or 8x GPUs	$2.63/hour	-	Single and multi-node cluster
HGX B200	TBA	$4.13/hour	-	Blackwell GPU architecture
GB200	36 Grace CPUs + 72 Blackwell GPUs	$8.00/hour	-	Liquid-cooled rack
L40S	2x, 4x, 8x GPUs	$1.80/hour	48GB	PCIe configuration
RTX PRO 6000	1-8x GPUs	$1.24/hour	-	Blackwell Architecture
RTX A6000	1x or 2x GPUs	$0.60/hour	48GB	PCIe configuration
RTX A5000	1x or 2x GPUs	$0.45/hour	24GB	PCIe configuration

Pricing is shown as starting rates per hour. Actual pricing may vary based on configuration, region, and availability.

Getting Started

To deploy GPU compute resources:

Choose your GPU configuration based on your workload requirements
Create a machine using the Airon CLI or API
Configure your environment with the necessary frameworks and libraries
Deploy your workload and monitor performance

# Example: Create a machine with 4x H100 GPUs
airon machines create \
  --type gpu \
  --architecture x86 \
  --brand nvidia \
  --model h100 \
  --number 4 \
  --region us-west-2

For more detailed guidance on GPU compute deployment, see our Getting Started guide or explore our tutorials.

Getting Started

Compute

Network

Inference

CLI

Plugins

Getting Started

​Getting Started

Getting Started