Skip to content

HPC Cluster Formations

Deploy dedicated High-Performance Computing clusters for scientific computing and simulations.

Overview

HPC Cluster formations provision dedicated High-Performance Computing clusters for workloads that require tightly coupled parallel processing, low-latency networking, and shared file systems.

Available in On-Demand and Spot modes.

Supported Technologies

Technology Description
AWS ParallelCluster AWS-native HPC cluster management
Terraform scripts Custom HPC deployments via IaC

Architecture

Component Description
Head Node Cluster controller running the job scheduler
Compute Nodes Scalable pool of compute instances for job execution
Shared Storage Shared file system accessible by all nodes

Configuration

Setting Description
Cloud Key Credentials for the target cloud account
Head Node Instance Type Compute resources for the head node
Compute Node Instance Type Compute resources for worker nodes
Compute Node Count Min/max/desired number of compute nodes
Shared Storage File system type and size
Job Scheduler Slurm or other supported scheduler
IaC Template CloudFormation, Terraform, or Pulumi script

Use Cases

Use Case Description
Scientific simulations CFD, molecular dynamics, climate modeling
Genomics Genome assembly, variant calling, population genetics
Financial modeling Monte Carlo simulations, risk analysis
Engineering FEA, structural analysis, crash simulations