Building a High Availability Azure Environment for HPC at Scale

Overview

A global semiconductor company known for its innovation in intelligent power and sensing solutions, supports some of the most demanding workloads in the automotive, industrial, and cloud sectors. As part of a broader cloud transformation initiative, Onsemi set its sights on modernizing its datacenter capabilities to meet the high availability (HA) demands of its HPC environments-particularly for its Pune, India facility.

This site, home to critical engineering workloads, lacked a resilient architecture capable of meeting the organization’s evolving uptime, replication, and scalability requirements. Our client’s team needed a solution that could not only establish a fault-tolerant HA infrastructure in Azure but also be easily replicated for future global deployments. A key challenge involved aligning Azure-native services with existing HPC practices while ensuring minimal disruption to engineering operations.

Solution

Oakwood Systems Group partnered with the client to deliver a repeatable, scalable high availability solution for their HPC environment, leveraging Microsoft Azure and infrastructure-as-code methodologies. The engagement began with a deep-dive technical assessment of the Pune datacenter’s current HPC methodology, including its use of Azure CycleCloud for workload orchestration, the configuration of compute clusters, storage topology, scheduler usage, and data locality policies.

With the goal of enabling a highly available architecture, Oakwood led a series of collaborative working sessions to document and dissect existing processes. These sessions enabled the team to design a resilient architecture that would serve as a production-grade HA replica of the Pune environment. Azure Site Recovery was architected as a key component of the solution, providing data replication and synchronization across regions, ensuring business continuity in the event of site failure or degradation.

To streamline the deployment and ensure repeatability, Oakwood developed a fully parameterized Terraform script designed to provision the entire HA environment in Azure. This script not only automates the deployment of networking, compute, storage, and recovery services, but also accommodates policy-driven customization for future environments. The HA deployment architecture was validated through extensive testing and quality assurance prior to production cutover.

Performance baselines were established and continuously tuned during this process. Throughout the implementation, Oakwood remained embedded with the client’s IT and engineering teams to provide optimization guidance, troubleshoot real-time issues, and adjust configurations to align with real-world usage patterns.

Outcome

The completion of this project resulted in a robust and repeatable high availability architecture that allows them to confidently run their HPC workloads in Azure with greatly improved resilience and disaster recovery capabilities. The Terraform automation developed by Oakwood now serves as a reusable asset for their global IT organization, enabling rapid deployment of future environments with a consistent, compliant, and fully documented approach.

By aligning Azure-native tooling with existing HPC processes, Oakwood helped eliminate configuration drift, reduce the time and risk associated with environment provisioning, and improve operational continuity. This project not only addressed an immediate need at the Pune datacenter but also established a scalable model for the client’s future cloud-first strategy – where infrastructure resilience and deployment velocity are non-negotiable.

About Oakwood

Our Team of seasoned professionals delivers unparalleled expertise in consulting and implementation services across the Microsoft Azure stack, ensuring our clients harness the full potential of their technology investments. With a commitment to excellence and a passion for driving business success, Oakwood Systems stands at the forefront of technological innovation, helping businesses navigate and excel in today’s fast-paced digital landscape.