Reference Architectures

CloudFormation templates and deployment guides for distributed training infrastructure.

🖥️

Compute

🖥️
Amazon EKS
Amazon EKS reference architecture for distributed AI on AWS
 🖥️
SageMaker HyperPod
Managed GPU clusters with resilient training and automatic recovery
 🖥️
AWS Parallel Computing Service
AWS Parallel Computing Service reference architecture for distributed AI on AWS
 🖥️
SageMaker HyperPod (EKS)
SageMaker HyperPod (EKS) reference architecture for distributed AI on AWS
 🖥️
SageMaker HyperPod (Slurm)
SageMaker HyperPod (Slurm) reference architecture for distributed AI on AWS
 🖥️
AWS ParallelCluster
AWS ParallelCluster reference architecture for distributed AI on AWS
🌐

Networking

🌐
VPC & Networking
EFA-optimized VPC configuration for high-bandwidth training
💾

Storage

💾
Common Resources (S3 & Events)
Shared S3 bucket and event notification setup
🛡️

Resilience

🛡️
NVRx
NVIDIA's resilient training toolkit for fault-tolerant distributed workloads