Case Studies
18 September, 2025
July 16, 2025
In a groundbreaking collaboration, TeraSky partnered with a premier scientific and engineering research institution to modernize their high-performance computing (HPC) and cloud-native application environments. This ambitious project not only involved a full-scale migration from Google Cloud Platform (GCP) to Microsoft Azure but also a complete overhaul of critical HPC workloads and Kubernetes-based application stacks, laying the foundation for enhanced agility, scalability, and innovation.
From GCP to Azure: A Strategic Cloud Shift
At the heart of this transformation was the migration of compute-intensive Weather Research and Forecasting (WRF) model simulations from GCP to Azure, along with a complete transition of containerized workloads from Google Kubernetes Engine (GKE) to Azure Kubernetes Service (AKS).
TeraSky led the effort to re-engineer all Helm charts, ensuring seamless workload portability and optimized performance on Azure. This involved deep cloud-native re-architecture that incorporated modern DevOps practices, infrastructure automation, and observability frameworks — all designed to accelerate research velocity and streamline operations.
Architecting a Scalable Azure HPC Environment
To meet HPC’s performance demands, Azure CycleCloud was chosen as the orchestration backbone. TeraSky deployed it to automate cluster provisioning, implement dynamic scaling, and enable intelligent job scheduling for simulation workloads.
Key infrastructure components were built using Terraform and Infrastructure-as-Code (IaC) principles. This included secure, high-bandwidth virtual networks, subnets, and compute pools optimized for inter-node communication — all essential for high-performance parallel processing.
In parallel, CI/CD pipelines were implemented using Azure DevOps, streamlining deployments, job validation, and version control. This DevOps framework also included automation for AMI image builds and data preprocessing scripts, speeding up model iterations and operational readiness.
Seamless Kubernetes Migration to AKS
TeraSky facilitated a strategic and disruption-free move from GKE to AKS by meticulously reverse-engineering and rebuilding the customer’s Helm charts. The result was a highly portable and cloud-native deployment architecture that optimized container orchestration for Azure’s ecosystem.
This transition enabled the customer to maintain agility, minimize service downtime, and improve the time-to-market for their core applications.
Visibility, Control, and Security at Scale
TeraSky implemented an advanced monitoring and observability suite using Azure Monitor, Grafana, and Loki. This enabled real-time telemetry, centralized logging, and proactive alerting across both HPC and AKS environments.
Financial governance was strengthened using Azure’s native cost analysis tools and a robust resource tagging strategy. This provided detailed cost attribution and dynamic resource scaling capabilities — critical for optimizing simulation schedules and budget usage.
Security was a core priority. TeraSky enforced Azure RBAC, implemented SSL certificate automation for CycleCloud, and followed strict tagging and isolation practices to ensure a hardened security posture and full policy compliance.
Business Impact: Quantifiable Innovation
This comprehensive modernization effort delivered major strategic and operational benefits:
Accelerated Research and Scientific Innovation
Cost Optimization and Financial Control
Operational Agility and Efficiency
Security and Governance Reinforced
A Future-Proof Cloud-Native Foundation
Ready to Modernize Your HPC Environment?
This project demonstrates the power of Microsoft Azure, combined with TeraSky’s deep expertise in HPC, Kubernetes, and DevOps, to drive innovation, efficiency, and cost control. Whether you’re looking to migrate, modernize, or build a future-ready cloud platform, TeraSky is your trusted partner for success.