HPC & CLOUD WHITEPAPERS

This page has been created to share whitepapers published by Red Oak Consulting.  Scroll down to view the various types of whitepapers and the topics our team of consultants have focused on.

Deploying, Benchmarking and Visualising workflows using an AMD HPC Cluster in Microsoft Azure

Once again Red Oak Consulting is proud to announce for the second year running, we were chosen to take part in the Microsoft Azure Centre of Excellence program.  This year, in partnership with Microsoft Azure and AMD our project was to demonstrate the performance and scalability of AMD EPYC™ based HPC virtual machines (VMs) in the Azure cloud using representative, readily available finite element analysis (FEA) and computational fluid dynamics (CFD) models.

Supporting this whitepaper are a series of three video tutorials recorded  by members of the Red Oak Consulting team.  The whitepaper is one document, with clearly defined sections.

Section one supports our first video which was recorded by our Managing Consultant, Nick Skingle.  In this section Nick explains how to deploy an AMD HPC cluster in Azure using automation via Azure DevOps.

Section two supports our second video which was recorded by our Principal Consultant, Manveer Munde.  In this section Manveer demonstrates the benchmarking of the AMD based HPC cluster using two softwares: Ansys LS-DYNA and OpenFOAM.  We also demonstrate the price and performance comparisons including Pay as you go (PAYG) instances and Spot instances.

And finally, section three which supports the final video, is where our Technical Consultant, James Page, demonstrates remote visualisation in Azure using and AMD GPU based remote desktop.

Click here to download the whitepaper.

How to Migrate HPC environments to Microsoft Azure

In 2020 Microsoft Azure launched its very first HPC & AI Collaboration Centre program.  The program was established to educate on specific HPC areas in collaboration with key Microsoft partners.  We are proud to say Red Oak Consulting is one of those partners.

Through the Microsoft Centre of Excellence program, supported by Intel, Red Oak Consulting produced three white papers on how to deploy HPC environments to the Microsoft Azure cloud service.

The three whitepapers focus on the whole journey: from the initial planning stages, right through to full deployment of a production-ready system.  Having completed an unrivalled number of HPC Cloud projects, the content of the whitepapers reflects our recommendations for best practice based on the lessons learnt from all the experience gained over the last few years.

Whilst the content of the whitepapers is applicable to any customer, in any industry, please note we have paid particular attention to the higher education and research-based community.  We focus on the Azure CycleCloud HPC orchestration and management service. CycleCloud, when setup appropriately, allows the creation of a familiar HPC system interface for users which mirrors that of a typical higher education based HPC environment.

In addition to three white papers, we have also produced three video tutorials allowing you to tangibly visualise the process on how to deploy a HPC environment on Microsoft Azure using Azure CycleCloud.

Whitepaper One: From requirement to proof of concept, preparing for HPC and research computing in the Azure Cloud

The first white paper considers the prerequisites and readiness requirements to get started with the cloud migration journey. Click here to download.

Whitepaper Two: Building and running a Proof of Concept in the Azure Cloud

The second white paper discusses how to create a simple proof of concept (PoC) HPC environment on Microsoft Azure to gauge the performance of HPC workflows in the Cloud. Click here to download.

Whitepaper Three: Considerations for the migration of a HPC environment to the Azure Cloud

This final white paper explores in detail important points to bear in mind when considering a production-level migration of a HPC environment to Microsoft Azure. Click here to download.