AWS G7e instance types
G7e is a gpu & accelerated Amazon EC2 family with 6 instance types, from 8 – 192 vCPUs and 64 GiB – 2 TiB of memory with NVIDIA RTX PRO 6000 Blackwell accelerators, running on Intel Xeon Scalable (Emerald Rapids). Linux on demand starts at $3.363 per hour.
gpu & acceleratedx86-64
- sizes
- 6
- vCPUs
- 8 – 192
- memory
- 64 GiB – 2 TiB
- accelerator
- RTX PRO 6000 Blackwell
- on-demand from
- $3.363
per hour · 13 regions
G7e sizes and prices
Linux, hourly, US East (N. Virginia) (other regions on each size's page)
| Size | vCPU | Memory | Accelerator | Local storage | On-demand | Monthly |
|---|---|---|---|---|---|---|
| g7e.2xlarge | 8 | 64 GiB | 1 × NVIDIA RTX PRO 6000 Blackwell | 1.9 TB NVME | $3.363 | $2,455 |
| g7e.4xlarge | 16 | 128 GiB | 1 × NVIDIA RTX PRO 6000 Blackwell | 1.9 TB NVME | $3.998 | $2,919 |
| g7e.8xlarge | 32 | 256 GiB | 1 × NVIDIA RTX PRO 6000 Blackwell | 1.9 TB NVME | $5.268 | $3,846 |
| g7e.12xlarge | 48 | 512 GiB | 2 × NVIDIA RTX PRO 6000 Blackwell | 3.8 TB NVME | $8.286 | $6,049 |
| g7e.24xlarge | 96 | 1 TiB | 4 × NVIDIA RTX PRO 6000 Blackwell | 7.6 TB NVME (2 disks) | $16.57 | $12,098 |
| g7e.48xlarge | 192 | 2 TiB | 8 × NVIDIA RTX PRO 6000 Blackwell | 15.2 TB NVME (4 disks) | $33.14 | $24,195 |
G7e equivalents on other clouds
closest match to g7e.8xlarge (32 vCPU, 256 GiB)
AzureNC40ads H100 v5NCadsH100v5 family · 40 vCPU · 320 GiB · 1 × NVIDIA H100 NVL$6.98/hr · East USNCadsH100v5 family →
Google Cloudg4-standard-48G4 family · 48 vCPU · 180 GiB · 1 × NVIDIA RTX PRO 6000 Blackwell$4.50/hr · IowaG4 family →
Other gpu & accelerated families on AWS
G7e FAQ
What sizes does the AWS G7e family have?
G7e has 6 instance types, from g7e.2xlarge (8 vCPU, 64 GiB) to g7e.48xlarge (192 vCPU, 2 TiB). They run on Intel Xeon Scalable (Emerald Rapids).
What is the cheapest G7e instance type?
g7e.2xlarge at $3.363 per hour (Linux, on demand) in US East (N. Virginia), about $2,455 per month.
What is G7e best for?
G7e is a gpu & accelerated family. GPUs and AI accelerators for training, inference, rendering and video.