GPU & Hardware to AWS Spot Instance Map
Quickly match GPUs and hardware to AWS EC2 Spot instance types. Ideal for sales calls or technical discussions when someone needs something specific—e.g. "I need H100" or "What Spot instances have A100?"
Perpetual Compute Supported (5 instance types)
These instance types are supported for Perpetual Compute and are available for deploy in the portal where listed. Pricing links go to a dedicated page when available, or otherwise to the general pricing page.
GPU → AWS Spot Instance Mapping
15 instance types shown
P5 (H100)
8× NVIDIA H100 (80GB HBM3)
LLM training · Large-scale inference · HPC · Scientific computing
P4d (A100)
8× NVIDIA A100 (40GB SXM4)
ML training · Deep learning · LLM inference · Distributed training
P4de (A100 80GB)
8× NVIDIA A100 (80GB)
Large model training · Memory-intensive ML
P3 (V100)
8× NVIDIA V100 (16GB)
Legacy ML training · Inference workloads
4× NVIDIA V100 (16GB)
Medium-scale training · Inference
1× NVIDIA V100 (16GB)
Small-scale training · Development
G6 (L4)
8× NVIDIA L4 (24GB)
Multi-GPU inference · Batch inference · Light training
G6e (L40S)
1× NVIDIA L40S (48GB)
Inference · Rendering · GenAI
G5 (A10G)
4× NVIDIA A10G (24GB)
Multi-GPU inference · Rendering
1× NVIDIA A10G (24GB)
Inference · Rendering · Virtual workstations
G4dn (T4)
4× NVIDIA T4 (16GB)
Inference · Batch processing
1× NVIDIA T4 (16GB)
Inference · Development · Cost-effective GPU
T3 (CPU)
No GPU — 2 vCPU, 4 GB RAM
Dev servers · Light workloads · CI/CD
Full reference table
| Instance Type | Hardware / GPU | vCPUs | Memory | Use Cases | Perpetual Compute |
|---|---|---|---|---|---|
| p5.48xlarge | 8× NVIDIA H100 (80GB HBM3) | 192 | 1920 GB | LLM training, Large-scale inference, HPC, Scientific computing | Supported |
| p4d.24xlarge | 8× NVIDIA A100 (40GB SXM4) | 96 | 1152 GB | ML training, Deep learning, LLM inference, Distributed training | — |
| p4de.24xlarge | 8× NVIDIA A100 (80GB) | 96 | 1152 GB | Large model training, Memory-intensive ML | — |
| p3.16xlarge | 8× NVIDIA V100 (16GB) | 64 | 488 GB | Legacy ML training, Inference workloads | — |
| p3.8xlarge | 4× NVIDIA V100 (16GB) | 32 | 244 GB | Medium-scale training, Inference | — |
| p3.2xlarge | 1× NVIDIA V100 (16GB) | 8 | 61 GB | Small-scale training, Development | — |
| g6.48xlarge | 8× NVIDIA L4 (24GB) | 192 | 768 GB | Multi-GPU inference, Batch inference, Light training | Supported |
| g6.2xlarge | 1× NVIDIA L4 (24GB) | 8 | 32 GB | Inference, Fine-tuning, Development | Supported |
| g6.xlarge | 1× NVIDIA L4 (24GB) | 4 | 16 GB | Inference, Prototyping, Small models | Supported |
| g6e.xlarge | 1× NVIDIA L40S (48GB) | 4 | 16 GB | Inference, Rendering, GenAI | — |
| g5.12xlarge | 4× NVIDIA A10G (24GB) | 48 | 192 GB | Multi-GPU inference, Rendering | — |
| g5.xlarge | 1× NVIDIA A10G (24GB) | 4 | 16 GB | Inference, Rendering, Virtual workstations | — |
| g4dn.12xlarge | 4× NVIDIA T4 (16GB) | 48 | 192 GB | Inference, Batch processing | — |
| g4dn.xlarge | 1× NVIDIA T4 (16GB) | 4 | 16 GB | Inference, Development, Cost-effective GPU | — |
| t3.medium | No GPU — 2 vCPU, 4 GB RAM | 2 | 4 GB | Dev servers, Light workloads, CI/CD | Supported |
Need a different instance type? Contact us to discuss adding support for your workload.
View live pricing