Manufacturing

AWS-Native HPC Migration Accelerates Engineering Simulations for a premium home appliances manufacturer

How Invisibl Cloud’s HPC migration strategy enabled engineering teams at a premium home appliances manufacturer to achieve 50% faster simulations, seamless scalability, and enhanced security by moving engineering simulations workloads to AWS HPC.

Faster CFD & Aero-Acoustic Workloads

Lower HPC Compute Costs

of Concurrent Simulations

The Challenge

The customer’s advanced R&D workloads required massive compute scale and stronger operational controls.

Limited Scalability

On-prem HPC clusters could not expand fast enough for increasing CFD, aero-acoustic, transient, & PowerFLOW simulation workloads.

Performance Bottlenecks

High-memory & CPU-heavy engineering jobs suffered delays, impacting product design cycles & extending experiment turnaround times.

Security & Compliance

Global R&D operations required a geographically distributed architecture with guardrails and governance, which did not exist across their HPC estate.

Lack of Observability

Monitoring job queues, resource utilization, & cost consumption was manual & reactive, causing inefficiencies & delayed insights.

Our Solution

End-to-end HPC migration and modernization using AWS ParallelCluster and AWS-native security, monitoring, and governance frameworks.

AWS Multi-Account Landing Zone Deployment

Set up a hardened multi-account architecture using AWS Control Tower, AWS Organizations, and Service Control Policies (SCPs), aligned with AWS Security Reference Architecture.

HPC Workload Migration to AWS

Migrated CFD DOE Optimization, aero-acoustic CFD, transient CFD, and PowerFLOW workloads to AWS HPC clusters orchestrated via Slurm on AWS ParallelCluster, with on-demand compute nodes and virtual workstations.

High-Performance Storage Architecture

Integrated Amazon FSx for Lustre for high-throughput scratch storage with seamless Amazon S3 connectivity for persistent data.

Monitoring, Reporting & Observability

Delivered Amazon CloudWatch dashboards and custom monitoring for CPU, memory, job performance, queue health, and cost visibility—with proactive alerts and automated reporting at user, queue, and cluster levels.

End-to-End Migration Support

Provided performance benchmarking pre-migration, and ongoing optimization, troubleshooting, patching, and cost governance post-migration.

Core Solution Features

Transformational Results

The HPC migration delivered measurable impact across performance, scalability, and governance.

Faster Simulations
Times

Parallel Jobs
With No Queue Delays

Lower
Compute Cost

Significantly Stronger
Security Posture

HPC Workflow Capabilities

Unified Job Execution

Run complex CFD and aero-acoustic simulations directly from optimized AWS HPC clusters using SLURM and virtual workstations.

Real-Time Monitoring

Track job concurrency, queue utilization, CPU/memory performance, and simulation health using dashboards and alerts.

Cost Management & Optimization

Monitor cluster spend, automate reports, and leverage spot instances and autoscaling to minimize compute costs.

Reporting & Analytics

Access automated reports covering cluster-level, queue-level, and user-level consumption patterns.

Fast Facts

Technologies

SLURM Job Scheduler

AWS EC2

Amazon S3

CloudWatch

GuardDuty

Security Hub

Amazon Inspector

AWS Control Tower

FSx Lustre

AWS Parallel Cluster

Dassault Systèmes PowerFLOW

CFD DOE

100+ Concurrent Workloads
Simulation job runtimes reduced from days to hours
Hardened security posture across global R&D teams
Fully automated monitoring & reporting
On-demand access to hundreds of compute cores