GPU Cloud Provider · Seattle, Washington, USA
Amazon Web Services
Amazon EC2 G4 instances provide cost-effective and versatile GPU options for machine learning inference and graphics-intensive applications using NVIDIA and AMD GPUs. These instances are optimized for various applications, including machine learning, gaming, and virtual workstations, providing a blend of performance and cost efficiency.
GPUs
15
Founded
Founded as part of Amazon Web Services, which was launched in 2006
GPU Marketplace
NVIDIA B200 SXMOn-Demand
NVIDIA B200 SXMOn-Demand
NVIDIA B200 SXMOn-Demand
NVIDIA B200 SXMOn-Demand
NVIDIA B200 SXMOn-Demand
NVIDIA B200 SXMOn-Demand
NVIDIA B200 SXMOn-Demand
NVIDIA B200 SXMOn-Demand
NVIDIA B300 SXMOn-Demand
NVIDIA B300 SXMOn-Demand
NVIDIA B300 SXMOn-Demand
NVIDIA B300 SXMOn-Demand
NVIDIA B300 SXMOn-Demand
NVIDIA B300 SXMOn-Demand
NVIDIA B300 SXMOn-Demand
Estimates only — rates are collected automatically from public provider pages and may be out of date. Prices vary by region, commitment term, and availability, and typically exclude storage, egress, and tax. Confirm current pricing with the provider before purchasing. Last collected .
Company Profile
FoundedFounded as part of Amazon Web Services, which was launched in 2006
HeadquartersSeattle, Washington, USA
Legal EntityAmazon Elastic Compute Cloud (Amazon EC2)
Infrastructure
GPU FleetNVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, up to 8 GPUs per G7e instance, 96 GB GDDR7 memory per GPU
Network FabricUp to 100 Gbps Ethernet, support for Elastic Fabric Adapter (EFA)
ConnectivityUp to 100 Gbps
StorageLocal NVMe-based SSD storage
Bare MetalDedicated Hosts (physical Amazon EC2 server fully dedicated for your use) are available.
AvailabilityGenerally Available
AI inference, scientific computing and spatial computing workloads
Compute & Deployment
On-DemandOn-Demand Instances offer pay-as-you-go compute capacity by the hour or second. No upfront payment or long-term commitment are required.
Spot / InterruptibleWith Amazon EC2 Spot Instances, you can use spare Amazon EC2 capacity in the AWS Cloud. This capacity is available at a discount of up to 90% compared to On-Demand prices.
Bare MetalA Dedicated Host is a physical Amazon EC2 server fully dedicated for your use. With Dedicated Hosts, you can use your existing server-bound software licenses. You can purchase them On-Demand (hourly) or as part of Savings Plans.
Container-BasedIf you prefer to manage your own containerized workloads through container orchestration services, you can deploy P6e-GB200 UltraServers and P6-B200 instances with Amazon Elastic Kubernetes Service(Amazon EKS) or Amazon Elastic Container Service(Amazon ECS).
Kubernetesyou can deploy P6e-GB200 UltraServers and P6-B200 instances with Amazon Elastic Kubernetes Service(Amazon EKS) or Amazon Elastic Container Service(Amazon ECS).
GPU Hardware
Latest GenNVIDIA RTX PRO 6000 Blackwell Server Edition GPUs
Multi-GPU Nodessupport NVIDIA GPUDirect RDMA for multi-GPU instances
Max GPUs/Node8
PCIe vs SXMNVIDIA GPUDirect P2P via PCIe
Pricing Model
Per HourOn-Demand Instances offer pay-as-you-go compute capacity by the hour or second.
SubscriptionOn-Demand Instances; Savings Plans; Spot Instances
Reserved Discountup to 72% compared to On-Demand prices
Spot Discountup to 90% compared to On-Demand prices
Pay-as-you-goOn-Demand Instances offer pay-as-you-go compute capacity by the hour or second. No upfront payment or long-term commitment are required.
Performance & Scaling
Multi-Node TrainingAdditionally, these instances offer 4x higher EFA networking bandwidth (1600 Gbps) compared to G6e instances and support NVIDIA GPUDirect RDMA for multi-GPU instances, enabling customers to use G7e instances for small-scale multi- node fine-tuning and training.
Elastic ScalingWith Amazon SageMaker HyperPod, G7e instances can scale to dozens of GPUs to train a model quickly without worrying about setting up and managing resilient training clusters.
Perf IsolationG7e instances are built on the AWS Nitro System, a combination of dedicated hardware and lightweight hypervisor which delivers practically all of the compute and memory resources of the host hardware to your instances for better overall performance.
Noisy NeighborThe AWS Nitro System with its specialized hardware and firmware designed to enforce restrictions so that no one, including anyone at AWS, can access your sensitive AI workloads and data.
Developer Experience
OnboardingSign up for an AWS account; Try Amazon EC2 with AWS Free Tier
FrameworksSupport for major machine learning frameworks compatible with NVIDIA and AMD GPUs
DocumentationGuides and tutorials; 10-minute tutorials; AWS Deep Learning AMIs; AWS Deep Learning Containers; product getting-started content
API FeaturesAWS CLI, SDK support for popular programming languages, REST API, AWS CloudFormation
Security & Compliance
SecurityRegular security assessments,Complies with AWS's comprehensive security protocols
ComplianceCompliance with industry standards as part of AWS's broad compliance portfolio
Customer testimonials (Agility RoboticsSynopsysARIBraveInnoactive)AWS Nitro System security claim
Data Center Locations
Coverage
all Regions and Availability Zones
Compliance Regions
Datacenter Locations
Key Strengths
High-performance G7e instances accelerated by NVIDIA RTX PRO 6000 Blackwell GPUs delivering high GPU memory (96 GB per GPU), up to 8 GPUs per instance, high memory and inter-GPU bandwidth, up to 1600 Gbps networking with EFA
Nitro System providing near-bare-metal performance and security, and up to 2.3x inference performance compared to prior generation (G6e).
Additional Information
Support Options
["24/7 support through AWS Support plans (Basic, Developer, Business, Enterprise)"]
Core Proposition
deliver cost-effective performance for generative AI inference workloads and the highest performance for spatial computing workloads
Notable Customers
Agility Robotics
Synopsys
ARI (Assured Robot Intelligence)
Brave
Innoactive
Last updated March 2026. Information subject to change.