Azure HPC CORE: 6-Week PoC

Oakwood’s Azure HPC CORE PoC provides a cost-effective, low-risk entry into high-performance computing (HPC) on Azure, ideal for organizations with lightweight workloads. By leveraging Azure’s flexible, cloud-based infrastructure, this PoC enables clients to explore HPC without the high upfront costs and complexities associated with on-premises systems.

With Slurm configured for efficient job scheduling and Azure Blob storage for seamless data management, this PoC will allow users to assess Azure's scalability and reliability for small-scale HPC tasks. Azure’s pay-as-you-go model ensures that organizations only pay for the resources they need, offering a low-cost option to explore HPC capabilities and scale as their requirements evolve.

Delivery Approach: • Assessment: Determine specific workload needs and objectives. • Design: Architect a minimalistic setup to handle low-resource HPC tasks. • Deployment: Set up CycleCloud, Slurm, and basic Azure VMs for efficient job management. • Testing: Execute introductory-level jobs to gather performance data. • Review and Report: Summarize key findings and provide scalability insights.

Deliverables: • Basic CycleCloud cluster deployment with Slurm for job scheduling. • Setup of VMs and Azure Blob storage. • Testing and analysis of lightweight HPC workloads on Azure.

Notes: • In addition to the 6 weeks, the Oakwood Team will devote an additional 2 weeks to work with the organization/user(s) to further optimize and manage the environment to ensure optimal performance. • The organization/user(s) will be limited by the number of jobs that can be ran during this PoC. • At this point a discussion will be had around moving forward with a full production environment or decommissioning the infrastructure that was stood up during the PoC.

Is Oakwood’s Azure HPC CORE PoC right for you? • Targeted Use: Lightweight workloads. • Complexity & Level of Effort: Low complexity and minimal setup. This PoC involves basic Slurm configuration and minimal infrastructure overhead, with a focus on simple job scheduling and cost-effective resources. • Scalability: Limited scalability, as it’s designed primarily for introductory-level HPC tasks.

Recommendation Guide: • If you need: A low-cost, low-risk environment to test Azure for HPC or evaluate lightweight workloads, the HPC CORE PoC is ideal. • Consider HPC PRO or MAX if: Your workload requires advanced storage solutions, mid- to high-level job scheduling, or higher performance needs.